There’s somebody on your team who manages a spreadsheet that’s core to your operation. Maybe it’s your bookkeeper closing out the month, your dispatcher reconciling the call ledger, your project coordinator patching the job-costing sheet. Maybe it’s you, running the analysis on your next big contract or your next big market move.

What is the spreadsheet for?

The pipeline tracker isn’t the point — the call to unstick the critical deal is. The job-costing sheet isn’t the point — the site walk to fix the estimate is. The cash-flow forecast isn’t the point — knowing when to push receivables is. The spreadsheet is the means. The action you take as a result — the call, the walk, the decision — is the end.

But it takes work — informational work — to get to the end.

AI is the next frontier of informational work. You or your team might not need to wrangle that spreadsheet quite so hard before long. You could have an AI agent for that. Which means you’d get that time back — the scarcest resource — to focus on the thing the spreadsheet is for. How you choose to reinvest that resource is the key question to wrestle with.

History rhymes, it doesn’t repeat

Spreadsheets themselves represent a historical example of this displacement and reinvestment dynamic.

In October 1979, Dan Bricklin and Bob Frankston shipped VisiCalc on the Apple II. Bricklin had been watching a Harvard Business School professor erase and rewrite a blackboard ledger every time a number changed, and he figured there had to be a better way. Turns out there was. Stores started selling the Apple II computer as a “VisiCalc accessory”. More than a quarter of every Apple II sold in 1979 went out the door to run it.

The thing the spreadsheet replaced wasn’t the thinking part of the work. It was the rote arithmetic underneath it — hand calculation, ledger posting, manual reconciliation. Believe it or not, “Computer” was an actual job title that humans held in 1979, spending all day literally doing arithmetic by hand, aggregating the results, and passing them back to a central synthesizer and organizer. I assume this was also a great time to be in orthopedics, specializing in carpal tunnel.

You might expect that this would reduce the number of bookkeepers needed across the economy — after all, much of the handiwork could now be done by digital computers, not living, breathing ones. That’s true, but it’s only half the story.

Since 1980, US bookkeeping and accounting clerk roles have shrunk by about 400,000 — but accounting jobs have grown by about 600,000. That’s a net positive. Economic historians call this creative destruction. When a new innovation arrives, it has displacing effects on the current state of work in that domain — and at the same time it opens up entirely new opportunities that couldn’t be contemplated before.

What happened? The work moved up the stack. People who’d been ledger clerks became analysts; analysts became Financial Planning & Analysis leads, business-intelligence partners, finance directors. Steven Levy named it in 1984, writing for Harper’s:

“Making spreadsheets, however necessary, was a dull chore best left to accountants, junior analysts, or secretaries… [now] senior executives who take the time to learn how to use spreadsheets are no longer forced to rely on their subordinates for information.”

— Steven Levy, A Spreadsheet Way of Knowledge, November 1984

But the spreadsheet revolution did something else, too. It didn’t just automate calculation. It made it cheap to capture, manage, and analyze entirely new information about the work — information that used to be too expensive to collect, store, and update. Shoshana Zuboff named this in her 1988 book In the Age of the Smart Machine, when she watched computers entering factories and offices. Computers didn’t only automate the work; they informated it - that is, created the need and means to capture and store previously unrecorded information. The work order. The dispatch log. The patient chart. The barcoded inventory. The customer-history note. All these now had a reason to become hard-coded into a hard drive somewhere - because they could be captured, analyzed, and stored so cheaply - digitally and not by hand.

Forty years of falling cost-to-capture-information is why even the most hands-on jobs have grown an administrative layer around them — because each piece of captured information was worth its declining cost.

AI is the next major chapter in that arc. Not the first one. Just the most aggressive yet. It’s compressing the cost of the information-shuffling layer your business sits inside — the administrative, analytical, coordinating wrapper around whatever your core business actually is.

So what’s the point from that quick history lesson? That revolution impacted your business whether you see it that way or not. The next one is coming for the same slice, faster and harder. How do we interpret that change, and turn that reality into a thoughtful decision rather than just something to cope with?

A framework: Three types of work

Every business is engaged in three broad types of work. Not jobs — types of work, mixed in different ratios across every role.

Informational work. Reading, writing, calculating, looking up, building the spreadsheet, building the deck, finding the answer. Shorthand: if you’re on a computer and there’s no human directly on the other side, it’s informational work.

Physical work. Installing the system, repairing the equipment, walking the jobsite, driving the truck, treating the patient. Shorthand: if you’re moving physical objects, it’s physical work.

Relational work. Talking to the customer, mentoring the new hire, negotiating with the supplier, developing a partnership. Shorthand: if you’re in a live one-to-one with another human, it’s relational work.

Every role on your team sits somewhere inside that triangle, and every role is a mix of all three layers, never just one vertex. Think of it as the role’s work fingerprint. The Technician sits low and left, mostly physical — but logs parts on a tablet (informational) and explains the findings to the homeowner or equipment manager (relational). The Operations Manager sits near the center because every day pulls them across all three — reviewing dashboards, walking the floor, coaching a tough conversation. The minimum slice of any one layer is almost never zero. AI gets in on the informational slice no matter how large or small.

So, do informational workers go away?

That’s the dominant narrative right now, particularly among the gloomier voices. Look at who’s writing it: academics, journalists, software engineers, consultants (hello!), podcasters. People whose work is almost entirely informational. Their sense of disruption is real — they’re watching their own jobs transform. But what comes in is a worldview error. They end up extrapolating that disruption to the whole economy, assuming most valuable work is fundamentally informational.

The numbers say otherwise. Anthropic, the company behind Claude, has their own Economic Index report from January 2026. Of 756 occupations they tracked, more than half show essentially zero Claude exposure — almost all of them physical trades, transportation, agriculture, food, and personal-care work. The BLS projects construction, healthcare delivery, and personal services to keep growing through 2034, not shrink. The categories that are mostly physical and relational keep growing despite all the AI doomerism.

This hits home for me personally at Conversint. I’ve been a technology consultant for most of my career. At Deloitte I spent much of my time researching, synthesizing, and building decks to carry an idea into a room — the kind of work AI hits hardest. Now, building Conversint, I feel like I have a team of highly paid consulting analysts at my fingertips. The result isn’t that I produce more decks. It’s that I can go deeper informationally faster, and show up (sometimes physically) better-prepared for the relational moments that matter — the conversations with people whose businesses we can actually help.

If your business is mostly informational — you run a SaaS company, a marketing agency, a professional-services firm — AI hits a larger share of your pie, and warnings of disruption apply to you more directly. Even then, the physical and relational layers of your business, may be the key to where your competitive edge may live. Or you double down on the informational — figure out how to use these tools to sling around more information faster and better than your competition.

For everybody else — the operator running a real-world business — AI hits the smaller part of the pie. Hard. The other two are largely untouched.

This dynamic brings us back to where we started: how do you reinvest the time you’re gaining back on the informational side?

Where does the time go? Augment to Expand vs. Automate to Extract

AI is effectively handing you time back to reinvest. You’re used to making reinvestment calls — does free cash flow go back into the business, or to the owners, or onto the balance sheet? AI is trying to hand you a new form of capital to allocate in terms of time saved on informational work. There are two main reinvestment paths to consider

The first path is Augment to Expand. Reinvest the time AI gives you back into the business. There are two ways to do this, and they can run at the same time:

• Go deeper on the informational side. Capture and analyze information you couldn’t afford to before. The customer-history note becomes a real history. The job-costing sheet captures dimensions you used to drop. Predictive maintenance and customer churn flags emerge from data you were already collecting but never had hours to analyze. Better decisions, faster, on more questions.

• Reinvest in the physical and relational layers. Your bookkeeper closes the books in half the time, and those hours go into checking with customers to review the numbers and see how their operation is doing. Your dispatcher routes calls faster and spends the extra time coaching the field techs based on patterns they see across calls. The foreman walks three more sites a week. The technician spends more time with the homeowner or equipment manager explaining what they found to help them feel reassured.

Same philosophy underneath both flavors: the time goes back into the business — to expand what you and your team can do, what you know, how intently you can serve your customers.

The second path stems from a different philosophy: automate to extract. Same productivity gain, different decision. Cut the role. Hold the output. Pocket the savings. Margin grows, market share even may as well. Headcount doesn’t. This is where the AI doomers can be shown to be onto something — the more business leaders make decisions in this direction, the more viscerally the job displacement effect will be felt.

Same technology, two philosophies. The difference isn’t in the tool. It’s in how you choose to use it.

But there is a third path. A recent Gartner survey found that AI saves sellers nearly five hours per week — but 72% of sales organizations fail to reinvest that time in higher-value activities. The default, in other words, is to lose the question entirely. Deferred capital allocation decisions result in deferred returns.

AI doesn’t decide where the time goes — you do. The hours it’s handing back are bigger than any previous informational tool, which makes the decision bigger too. Past tools handed you minutes a day; AI is handing you hours a week per person.

How you choose to reallocate that capital will shape what your business looks like a year from now.

Say it with me: what is this spreadsheet for?

The question we opened with works as a Monday-morning exercise. Thirty minutes, a notebook, two or three spreadsheets.

Pick a spreadsheet you and your team rely on every week — the one you dread maintaining, the one you always have open. Write one sentence answering the question. Not what it tracks — what it’s for. “Knowing which customers haven’t heard from us in ninety days so somebody can pick up the phone.” That’s your for talking. “Tracking customer activity” is a means, not an end.

Now write a second sentence: what would you and your team be doing more of if this spreadsheet maintained itself? “Two more customer visits a week.” “Calling the five oldest overdue invoices every Monday.” “The foreman walking three more jobsites.” Notice that almost every answer lands on the physical or relational layer of your business. That’s not a coincidence — that’s the triangle showing up in your own handwriting.

Do that for two or three of your spreadsheets. The pattern will tell you where the gap is widest, and where the reinvestment question lands first.

What it’s always been about

The spreadsheet was always a means. AI is too. The end has always been the work your business actually does in the real world — the physical work and the relational work.

Where the time goes is your call. Reinvest it — into going deeper on the informational side, or into the physical and relational sides of your business that customers actually pay you for — and your team grows into bigger, more interesting work. Extract it as margin, and you’re trading longer term compounding and expansion for shorter term gains. Doing nothing means ceding ground to your competitors who start down one path or another.

Which path will you take?

This post was collaboratively brainstormed, researched, and drafted alongside AI (Claude Opus is our go-to lately for this sort of thing), then revised and edited heavily by real humans.

Conversint is in the business of augmenting you and your workforce. If you want to get serious about reinvesting the time AI gives you back, on whichever layer of your business it lands, reach out.

If you learned something useful, consider sharing this post with someone else who may find it useful as well.

Share

For more insights and guidance like this delivered to your inbox regularly, consider subscribing.

Subscribe now

Research Sources

1. Anthropic. (January 2026). Anthropic Economic Index — January 2026 Report. https://www.anthropic.com/research/anthropic-economic-index-january-2026-report

2. US Bureau of Labor Statistics. Occupational Outlook Handbook. https://www.bls.gov/ooh/

3. Zuboff, S. (1988). In the Age of the Smart Machine: The Future of Work and Power. Basic Books. https://www.basicbooks.com/titles/shoshana-zuboff/in-the-age-of-the-smart-machine/9780465032181/

4. Goldstein, J. & Kestenbaum, D. (2015). Planet Money episode 606: Spreadsheets! NPR. https://www.npr.org/sections/money/2017/05/17/528807590/episode-606-spreadsheets

5. Levy, S. (1984). A Spreadsheet Way of Knowledge. Harper’s Magazine, November 1984. https://harpers.org/archive/1984/11/a-spreadsheet-way-of-knowledge/

6. VisiCalc. Wikipedia. https://en.wikipedia.org/wiki/VisiCalc

7. Computer (occupation). Wikipedia. https://en.wikipedia.org/wiki/Computer_(occupation)

8. Gartner. (May 2026). Survey Finds AI Saves Sellers Nearly Five Hours per Week, Yet 72% of Sales Organizations Fail to Reinvest Time in High-Value Activities. https://www.gartner.com/en/newsroom/press-releases/2026-05-19-gartner-survey-finds-ai-saves-sellers-nearly-five-hours-per-week-yet-seventy-two-percent-of-sales-organizations-fail-to-reinvest-time-in-high-value-activities

Here’s something strange. If you were to survey your customers and ask them how they feel about AI in customer service, most of them will tell you to keep it away from them. Gartner put the question to 5,728 people and found that 64% would rather companies didn’t use AI for service at all. In fact, more than half said they’d switch to a competitor over it.

Now ask those same customers how they spend their Tuesday afternoon. They’re chatting with an AI, happily, to reschedule a flight, check on a package, or summarize Amazon reviews on some new tech they’ve been scoping. Zendesk’s research pegs the share of consumers actively wanting to use AI assistants for their own service tasks at 67%.

The contradiction resolves the moment you stop treating “AI” as one thing. Customers don’t hate AI. They hate AI deployed at them instead of for them. They hate being stuck, and putting a robotic bouncer that is having trouble interpreting their request adds insult to injury.

Here’s the challenging reality: a good AI interaction is a shrug; a bad one is an exit. Seventy percent of consumers say a single bad AI experience is enough to send them shopping for a new brand. A single great one? Barely moves the loyalty needle. The gains are linear. The losses are off a cliff.

Think of AI as a shortcut up the mountain. It’s a real shortcut — real time saved that can be reinvested — but only if you take the time to build the trail carefully. If you don’t, your customers don’t get a mildly inconvenient Tuesday. They go missing.

Depending on what business you’re in, you might be thinking: my customers are younger, more tech-savvy, they’ll love it. The data has a twist for you. Younger customers are more willing to use AI and more willing to walk when it fails. Gen Z and Millennials are enthusiastic adopters — as long as they can escape to a human. For younger generations, your web experience is your storefront. The experience they have at the digital doorstep says more to them about your brand and the value you place on them than you might think. The asymmetry is amplified.

The view from the top

Before we talk about the pitfalls — and there are a few — let’s be clear about why anyone would try to summit this treacherous mountain in the first place.

A well-built AI stack at a 10-to-100-person company will reclaim roughly 10 to 20 hours a week per user. That’s not marginal. That’s a part-time employee, freed up, pulled off your most expensive headcount. The math is compelling.

Here’s the fork in the trail. Most companies, when those hours come back, pocket them as margin and reduce headcount. Run a tighter ship. Do more with less. It looks fine on a spreadsheet — right up until a competitor decides to reinvest their reclaimed hours strategically.

The alternative is to focus those hours into the moments your enterprise-scale rivals cannot possibly replicate. For example:

First purchase. A hand-signed note in the box. A short welcome video. A founder call within the first week. Amazon can’t do this. You can.

Cancellation attempt. Not a form, not a survey. A real human on the phone within the hour, authorized to listen and authorized to fix. Winning a customer back is a fraction of the cost of finding a new one — and it’s effectively impossible to automate well.

Complaint recovery. A real person acknowledging the problem within the day, with the authority to make it right. This isn’t where loyalty is tested. It’s where loyalty is built.

Onboarding. A 15-minute call in the first week that nobody asked for. Your largest competitor cannot make this call at scale. You can make it effortlessly.

The annual check-in. Once a year, a human reaches out to every meaningful customer — not to sell them anything, just to check in. The cumulative effect over three years is enormous.

In a February 2026 HBR piece, Horst Schulze (the Ritz-Carlton cofounder) and Micah Solomon argue the same thing from the enterprise side: as AI becomes commoditized, human hospitality becomes the moat. Your enterprise competitors are spending billions trying to scale intimacy — trying to make AI feel human. You don’t have that problem. Your team already is human. The game isn’t scaling intimacy — it’s creating the time to deploy it where it matters. AI is how you fund that.

AI handles the routine so your humans can handle the signature. The question isn’t how much of your support to automate. It’s which stretch each interaction lives on.

Three stretches of trail

So where does AI actually belong on your trail map? Think of it in terms of terrain.

The green zones are the flat, paved parts of the trail. Routine, deterministic, low-emotion. Order status. Returns. Password resets. Scheduling. When a customer asks when their package arrives, they want the date. They don’t want a relationship. They don’t want empathy. They want the date, instantly, at 11 p.m., in whatever language they prefer. AI is genuinely better here — not because it’s cheaper, but because it’s faster.

The yellow zones are the rough patches where the trail is overgrown or worn out. You may need a guide to show you where to step, and it needs to be carefully maintained. These are the oddities. A billing question with one eyebrow-raising detail. A return that falls in a gray zone. This is the copilot zone — AI drafts, a human reviews. AI summarizes, a human decides. Done well, this doesn’t replace anyone. It sharpens them.

The red zones are the parts where you hand your gear to the Sherpa and let the expert lead the way. This is a full handoff from the AI to the rep, but with AI-enabled tooling supporting them in the background — retrieving specific policy wording, pulling up the customer’s history, drafting a response for the rep to approve. Cancellations. Complaints. Anything emotionally loaded. The conversations that decide whether the relationship survives. When a customer starts explaining why they want to cancel, they’re not really asking for a refund — they’re asking whether you’re the kind of company worth sticking with. No amount of clever prompting makes an AI the right voice for that.

Climbing without handrails

A few climbers have already fallen off this mountain. Worth a glance before you lace up.

Air Canada, February 2024: a grieving customer asked the airline’s chatbot about bereavement fares. The bot invented a policy that didn’t exist. When the airline refused to honor it, a Canadian tribunal ordered Air Canada to pay up and flatly rejected the argument that the chatbot was “a separate legal entity.” Your bot’s hallucinations are your hallucinations. That’s a wake-up call.

Around the same time, Klarna — the buy-now-pay-later fintech — announced its AI assistant was doing the work of 700 customer service agents. Fifteen months later, CEO Sebastian Siemiatkowski quietly reversed course and started rehiring humans. His own words: “We focused too much on efficiency and cost. The result was lower quality — and that’s not sustainable.”

And DPD, the parcel carrier, had a customer coax its chatbot into writing a poem about what a terrible delivery firm DPD was. Screenshots hit 1.3 million views in a single day.

The technology worked in all three cases. What failed was intent. One was optimizing for deflection. One for cost. One for volume. None of them was optimizing for the customer in front of the screen — and customers read that signal almost instantly.

Five handrails to build before you open the trail

If you’re building the trail, build these handrails first. You can skip any of them. You’ll regret every one you skip.

1. Start with deflection, not replacement. Target the routine 60–80% of tickets. Keep humans as the default for anything emotional, regulated, or relationship-defining. This is the biggest philosophical fork in the road. Replacement logic is what sent the Klarnas of the world scrambling to rehire. Deflection logic quietly delivers the same savings without breaking the brand. The vendor landscape has largely caught up with this framing — Intercom Fin, Zendesk AI Agents, Ada, Decagon, and Sierra all position around resolution of well-scoped tickets with a clear human fallback, rather than “replace your team.”

2. Invest in the knowledge base before the bot. Audit your existing help docs, old tickets, and support macros. Consolidate. Clean. Make it retrievable. A messy knowledge base doesn’t just limit AI — it amplifies your weak spots. The bot will confidently say the wrong thing, in your voice, faster than any human ever could. Research on retrieval-grounded AI is unambiguous: the knowledge base is the ceiling.

Good news — this is also the step where AI is most useful to you. Export a year’s worth of support tickets to a CSV and hand it to Claude Projects, ChatGPT with Advanced Data Analysis, or Google’s NotebookLM. Ask it to cluster the top 20 recurring themes, flag the questions your docs don’t answer well, and rank them by volume. You’ll get a prioritized work list for your content team in an afternoon — work that used to be a month of manual tagging. This is AI as a tool for you, before you point it at your customers.

3. Design the escalation path first, not last. Most of what customers call “a bad AI experience” isn’t about a failed answer. It’s about a failed escape. They asked for a human and couldn’t find one. Define your triggers (low confidence, emotion detected, specific keywords), define how the conversation context transfers to the human, and define the SLA for the follow-up. If the escape hatch is clean, customers will forgive a lot. Most major platforms now expose handoff logic and context-passing out of the box — use them, and test them with a stopwatch.

4. Match pricing to volume shape. Stable, predictable volume? Per-seat pricing is probably fine. Seasonal or spiky? Outcome-based pricing — roughly a dollar per resolved conversation, depending on vendor — finally aligns your vendor’s incentives with yours in a way per-seat never did. Re-evaluate at every 2× growth milestone. The alignment can flip on you fast if your volume changes shape.

5. Protect the signature moments. Identify the three to five interactions that define your brand — the ones a customer would tell their friend about. Route them to humans, enhanced by AI, never replaced by it. These moments carry wildly disproportionate weight in whether a customer stays.

Better questions, better directions

The wrong question is the one everyone asks first: how much of my support can I automate? It points you at the trail. It measures success in tickets deflected and dollars saved.

The right question is quieter: which interactions define my brand, and how do I use AI to give my humans more time and sharper tools to nail those? It points you at the summit.

If you want a low-friction place to start, try this. Open ChatGPT, Claude, or Gemini and paste in the following:

Act as a customer experience strategist interviewing me. I run a [describe your business in one sentence]. I want to identify the three to five “signature moments” in my customer journey — the interactions that most shape whether someone stays a customer, recommends us, or leaves. Ask me one question at a time. Start with the customer’s first contact and walk me through the relationship chronologically. After about 10 questions, summarize what you’ve heard, propose my top signature moments, and flag which ones a competitor three times my size couldn’t replicate.

Spend 20 minutes on that conversation. Then do a second pass with your last 90 days of support emails or tickets pasted in, and ask: “Where in this data do you see the most friction, confusion, or repetition? Cluster it by theme and estimate how many of these could be handled by a well-built AI response, versus which ones signal a signature moment I’d want a human to own.”

Those two conversations, back-to-back, will give you a sharper map of your own trail than most consulting engagements produce in a month.

The companies still asking “how much can I automate” are planning their own headcount reversal. The ones asking “what do I want my humans free to do” are building something enterprise-scale rivals cannot replicate — and quietly capturing the savings anyway. The shortcut was never the point. The time at the top was.

This post was collaboratively brainstormed, researched, and drafted alongside AI (Claude Opus is our go-to lately for this sort of thing), then revised and edited heavily by real humans.

---

Conversint is a consulting firm helping SMBs and everyday humans unlock real value from AI and other modern tech. To learn more about our services and schedule a free initial consultation, visit our website at Conversint.ai.

If you learned something useful, consider sharing this post with someone else who may find it useful as well.

Share

For more insights and guidance like this delivered to your inbox regularly, consider subscribing.

Subscribe now

At this point you probably have a few AI tools at your fingertips. If you’re involved in Sales & marketing, or you’re trying to grow your business, one intuitive application of them may be to streamline your cold outreach. More emails and messages, faster, to more people. This is the fast food model of lead generation — crank up the volume, keep the cost low, hope the numbers work out.

Here’s the problem: everyone had that same idea at the same time, and traditional channels are collapsing under the weight of it.

Google, Yahoo, and Microsoft have rewritten the rules for bulk email. Gmail now rejects non-compliant sends outright as of late 2025. Microsoft followed suit, enforcing its own sender requirements starting in May 2025. Spam complaint thresholds are razor-thin. And the numbers reflect the squeeze — B2B cold email reply rates have dropped from roughly 7% to around 4–5% in just two years. The channel itself is actually degrading because everyone automated it.

And if you’ve seen the tools that promise “personalized AI emails at scale” — inserting a first name and company into 2-3 AI-generated templates doesn’t fool anyone. It definitely doesn’t fool Gmail’s filters.

Here’s another perspective: the real value of AI in lead generation isn’t automating the customer-facing interaction — the dining room, if you will. It’s transforming the preparation that happens before any conversation takes place — the research, the strategy, the context-building. The kitchen work. Stop automating the dining room. Invest in a better kitchen.

The rest of this article shows you what that AI-enabled kitchen looks like at every stage — from understanding your market all the way down to the post-meeting follow up with an individual.

A Taste of What’s Possible

You just walked out of a prospect meeting. You were fully present the entire time — no scribbling notes, no splitting attention between listening and documenting. You picked up on hesitation when pricing came up. You noticed the technical buyer lean in when you mentioned implementation timelines.

Ten minutes later, a follow-up email is in your inbox for review. It references specific things discussed, captures the action items, and drafts next steps. You adjust the tone based on what you sensed in the room, add a line about implementation since that’s where the energy was, and hit send. Five minutes instead of thirty. If you don’t hear back, your AI-assisted prospecting tool will send automated nudges, while you move on to prep for the next potential client.

All this didn’t start when the meeting began, and sat down at the table with your prospect. It started with how you found that prospect, how you decided they were worth your time, and how you walked in already knowing what mattered to them.

The kitchen was working long before that. Let’s look at how.

Sourcing the Right Ingredients (Market-Level Strategy)

Before you research any specific company or prep for any meeting, you need to know where your audience actually is and what they care about right now.

AI helps you survey the full landscape: trending topics in your target market, channels your audience frequents, what competitors are publishing, and — most usefully — where the gaps are.

This is the level that feeds your content strategy and channel selection. It answers “where should I show up?” and “what should I be talking about?” — questions most small businesses answer with gut instinct or imitation.

It’s also where leverage lives in the form of multi-channel content. Once you’ve identified the right message and the right channels, AI helps you adapt a single piece of content across formats — a blog post turns into LinkedIn posts, an email newsletter, a short video script — without starting from scratch each time.

Think of this as intelligent lead sourcing. You’re going to the market with a plan, understanding what’s in season and what your guests actually want to eat. You’re not grabbing whatever’s on the shelf and hoping it works.

There are a few tools in the landscape for you to consider at this stage.

Start simple: Use Perplexity or ChatGPT with browsing to synthesize trends and run competitive analysis. Free, and surprisingly effective for a first pass.

Step up: SparkToro gives you real audience intelligence — where your target audience spends time online, who they follow, what they engage with. There’s a free tier with five searches a month, and paid plans start around $50/month.

For content adaptation: Castmagic turns long-form audio and video into multi-format assets for about $23/month, or you can use Claude or ChatGPT to reformat a blog post for different channels at no cost.

What AI does here: Surfaces trends, synthesizes competitive landscapes, and adapts content across formats and channels — fast.

What you do: Decide whether a trend is worth pursuing, whether your services can credibly address it, and whether what you’d publish is differentiated enough to matter — or just adding to the noise.

Prepping the Station (Account-Level Research)

Market intelligence tells you where to focus. Now you decide which specific companies warrant your attention.

AI helps build richer account profiles faster: what a target company has been in the news for, what their recent priorities signal, what roles they’re hiring for (which often reveals pain points), what their tech stack looks like.

The real value isn’t the raw data — it’s the pattern recognition it enables. AI can flag that a company just posted three data engineering roles. You recognize that means they’re probably six months from needing the kind of work you do, and you decide to nurture rather than hard-pitch. That timing instinct is yours, and more sophisticated AI systems can pick up on those cues based on your interaction patterns over time.

For smaller teams and solo operators, this is where AI levels the playing field. You’re building the kind of account intelligence that well-staffed enterprise sales teams take for granted — without the headcount.

Mise en place (everything in its place) — that’s the kitchen concept here. You’re not scrambling mid-shift to figure out what you’re working with. The station is prepped, and when it’s time to plate, you’re executing with confidence.

Start simple: ChatGPT or Claude with web search. Research a target company’s news, job postings, and recent developments, then ask it to synthesize signals and likely priorities. Free.

Step up: Apollo.io combines a contact database with engagement tracking. There’s a free tier, with paid plans starting around $49/month.

Go deeper: Clay automates multi-source enrichment and AI-driven research workflows, pulling from 130+ data providers. It starts around $149/month and is built for more technical users — think of it as the professional kitchen upgrade you invest in once you’ve validated the approach manually and are ready to systematize it. At the enterprise end, LinkedIn Sales Navigator and ZoomInfo serve a similar function at larger scale.

What AI does here: Assembles account profiles, flags hiring signals, surfaces news and priority shifts — breadth and speed across dozens of targets at once.

What you do: Decide which of those accounts actually match what you solve. Prioritize based on relationship proximity, strategic fit, and timing — judgment that combines market knowledge with instinct.

Personalizing the Dining Experience (Individual Lead Prep)

You’ve identified the right markets, prioritized the right accounts. Now you’re meeting a specific person.

AI assembles the context package in minutes: their recent LinkedIn activity, their company’s latest news, conference talks they’ve given, perspectives they’ve published, patterns in what they engage with.

This isn’t surveillance — it’s showing up informed enough to have a relevant conversation instead of a generic pitch. The difference between “tell me about your business” and “I saw your team just expanded into the Southeast — how’s that going?”

The simplest version requires no specialized tool and almost no learning curve.

Think of it as the pre-shift briefing. The hostess tells you the couple at table 4 is celebrating an anniversary, the regular at the bar just landed a big contract. You walk in knowing something that transforms a transactional interaction into a personal one.

Start simple: Export your prospect’s LinkedIn profile as a PDF (LinkedIn has this built in), upload it to Claude or ChatGPT along with their company’s about page or recent news, and ask for a meeting prep brief — conversation starters, likely pain points, themes from their recent activity. Five minutes, zero cost. The AI doesn’t find the information — you do. It synthesizes and connects it into something actionable.

Step up: LinkedIn Sales Navigator gives you deeper activity tracking, saved lead alerts, and relationship mapping for about $100/month.

Go deeper: Cirrus Insight delivers automated daily pre-meeting briefs straight to your inbox, or Clay can systematize prospect briefings assembled from multiple sources.

What AI does here: Synthesizes a person’s public activity, company context, and engagement patterns into a briefing you can act on — in minutes instead of an hour.

What you do: Read between the lines. You see the prospect recently posted about vendor consolidation frustration and decide to lead with empathy rather than your standard pitch. You notice they commented positively on a competitor’s case study and shift your positioning accordingly.

AI gave you the ingredients — now you’re cooking in real-time based on the dynamics of an actual conversation. Fully present. Which brings us back to where we started.

Back to the Table (Meeting Synthesis and Follow Up)

Remember that scene — the meeting where you were fully present, the follow-up that practically wrote itself? Now you can see the whole chain behind it.

Market intelligence shaped your content strategy. Account research told you this was the right company at the right time. Individual prep meant you walked in knowing what mattered to this specific person. And AI captured everything during the meeting so you could focus on the human sitting across from you.

Start simple: Built-in AI summaries in Zoom, Google Meet, and Teams. Free, already available, probably already turned on.

Step up: Fathom offers free unlimited recording with instant summaries and action items.

Go deeper: Fireflies.ai for CRM integration and cross-meeting analytics, or Otter.ai for real-time collaborative transcription.

What AI does here: Records, transcribes, summarizes, and drafts follow-ups so nothing falls through the cracks.

What you do: Stay present in the conversation. Read the room. Notice the hesitation, the enthusiasm, the unspoken concern. Then shape the follow-up based on what you sensed — not just what was said.

The Dining Room is Still Yours

Here’s the thread across all four levels: AI handles breadth and speed. You handle depth and judgment. As you move from market research down to the meeting itself, AI’s contribution gets more specific and your contribution gets more relational. Neither side is dispensable.

These layers compound. Any one of them is useful on its own. Together, they create a preparation infrastructure that fundamentally changes the quality of every interaction.

The best restaurants have never been about serving more tables. They were about making every guest feel like the only one in the room. AI can’t do that on it’s own. But it can make sure you’re ready to.

Not sure where to start? Pick one layer. The meeting tools are the easiest first step — try one on your next call and see what changes when you’re fully focused on the human in front of you!

This post was collaboratively brainstormed, researched, and drafted alongside AI (Claude Opus is our go-to lately for this sort of thing), then revised and edited heavily by real humans.

Conversint is a consulting firm helping SMBs and everyday humans unlock real value from AI and other modern tech. To learn more about our services and schedule a free initial consultation, visit our website at Conversint.ai.

If you learned something useful, consider sharing this post with someone else who may find it useful as well.

Share

For more insights and guidance like this delivered to your inbox regularly, consider subscribing.

Subscribe now

A potential customer picks up their phone and asks ChatGPT: “What’s the best accountant in Denver for small businesses?” Three names appear in the response. Your business, Denver Tax & Co., isn’t one of them.

This scenario plays out thousands of times daily across every industry imaginable. While you’ve been perfecting your Google rankings, a parallel universe of business discovery has emerged—and most businesses are completely invisible to it.

Welcome to the conversation economy. The way people find businesses is fundamentally changing, and AI assistants aren’t just answering questions anymore—they’re making recommendations. The question isn’t whether this shift affects your business. It’s whether you’ll be part of the conversation or left out entirely.

There are currently many acronyms currently vying for the title of this new entrant into the marketing arena, and you may have heard a few of them already: AI SEO (AI search engine optimization), AEO (Answer Engine optimization), GEO (generative engine optimization) LLMO (large language model optimization), and, my personal favorite, LMAO (large model ai optimization).

Alright I made that last one up. But you almost bought it because that’s just the sea we’re swimming in.

From Ten Blue Links to Friendly Advice

For two decades, the formula was simple. Someone typed keywords into Google, scanned a list of ten blue links, clicked through to several websites, and eventually made a decision. The game was about ranking—getting your link as high as possible on that results page.

That world hasn’t disappeared. But a new one has emerged alongside it.

Today, a growing number of people skip the list entirely. Instead of typing keywords, they ask questions. Instead of clicking through multiple websites, they receive a synthesized answer—often with specific recommendations baked right in. They might never visit a website at all.

The difference is a bit like going to the bookstore and asking the librarian which shelf to check versus texting your well-read friend for a recommendation. The librarian hands you a list of options. Your friend tells you what they think you should read based on what they know about you, your goals, and your preferences.

AI is becoming that well-read friend for millions of people. And if your business isn’t part of what it “knows,” you’re not getting recommended.

The Winds, They are a-Changin’

Talk data to me for a minute.

According to Search Engine Land, zero-click searches now account for roughly 27% of all Google searches—and that number climbs even higher when Google’s AI Overviews appear. SEOworks reports that by mid-2025, AI Overviews (that summary you get at the top of the Google page most of the time now) were showing up in nearly half of all searches. Projections suggest they could appear in over 80% of informational queries soon.

The game has shifted. It’s no longer just about getting the click. It’s about getting your brand into the responses AI provides when customers ask about topics relevant to your business.

Is this a real shift or just hype? The data points in one direction. Adobe’s analysis of over 1 trillion visits to U.S. retail sites found that traffic from generative AI sources grew 4,700% year-over-year by July 2025. That’s not a typo—nearly 50x growth in a single year.

Pew Research Center reports in June 2025 that 34% of U.S. adults have used ChatGPT—roughly double the share from 2023—and among adults under 30, that figure jumps to 58%.

This isn’t early-adopter territory anymore. It’s approaching mainstream.

The consumer behavior shift is equally striking. Adobe found that 38% of U.S. consumers have used generative AI for online shopping, with 52% planning to do so this year. And perhaps most importantly, this traffic converts. Webflow reports that their ChatGPT traffic converts at six times the rate of Google traffic. AI-referred visitors aren’t just browsing—they’re buying.

Now, a critical caveat: AI referral traffic still represents a small fraction of total web traffic for most businesses—often less than 1%. But that small slice is growing explosively, and the visitors it delivers are disproportionately valuable.

The question isn’t whether AI search matters today. It’s whether you’ll be visible once it reaches full mainstream adoption.

Should You Care? A Compass of Consideration

Not every business faces the same level of exposure to this shift. Think of it like waves approaching a shoreline—some businesses are already learning to surf, others are digging in to brace for impact, and some are just barely starting to notice the tide change.

Two main factors drive how much this affects you right now, for you to use as a compass of sorts:

Demographics: How much of your customer base is actually using AI tools? Younger consumers, higher-income brackets, and certain demographics show significantly higher usage rates. If your customers skew under 40 or earn over $100k, they’re more likely already asking AI for recommendations.

Research intensity: Does your business involve long purchasing cycles or complex decision-making? The more research a purchase requires, the more likely AI is already part of that journey.

Learning to surf

Travel and hospitality businesses are already feeling the impact—AI-driven traffic to travel websites has surged dramatically. Professional services face significant exposure too, particularly financial advisors, consultants, and legal professionals. One stat that stopped me in my tracks: 94% of B2B buyers now use AI tools during their buying journey. If you’re in these categories, the wave has arrived.

Bracing for impact

Real estate, insurance, automotive research, and e-commerce show strong AI referral patterns. You can see it coming. Early movers in these categories will capture major value before the wave fully hits.

Rising tides

Senior care, childcare, wedding planning, and home services may seem like surprising candidates. But these involve exactly the kind of complex, high-stakes decisions where AI’s ability to synthesize information provides enormous value. The research intensity is high, even if the demographics haven’t fully caught up yet. The tide is rising. It’s just a matter of when.

Window of Opportunity

Here’s what makes the current moment so interesting: most businesses have essentially zero AI visibility. Competition for AI citations is virtually nonexistent in many categories.

While thousands of businesses fight tooth and nail for top Google rankings, the parallel competition for AI recommendations remains wide open. The businesses that establish themselves now will likely maintain significant advantages as AI search scales.

This mirrors the early days of search engine optimization. In 1999, businesses that recognized SEO’s importance gained advantages that lasted years. The same dynamic is playing out today with AI visibility—but the window is compressed. Things move faster now.

AI Visibility Audit: Your Monday Morning Exercise

Here’s a simple exercise for you on a Monday morning: Open ChatGPT, Perplexity, or Google’s AI Overview and ask the questions your potential customers are asking. If someone asked about your industry right now—your specific category of service, in your location—would your business be mentioned? Would the information be accurate? Would the characterization be favorable?

Even more simply: if you just asked

“What is [your business name here]?”

Would you even show up?

The conversation about your industry is happening right now, with or without you. The question isn’t whether to join the conversation. It’s how quickly you can get there.

This post was collaboratively brainstormed, researched, and drafted alongside AI (Claude Opus is our go-to lately for this sort of thing), then revised and edited heavily by real humans.

Conversint is a consulting firm helping SMBs and everyday humans unlock real value from AI and other modern tech. To learn more about our services and schedule a free initial consultation, visit our website at Conversint.ai.

If you learned something useful, consider sharing this post with someone else who may do the same.

Share

For more insights and guidance like this delivered to your inbox regularly, consider subscribing.

Subscribe now

You know the one. Someone reaches out asking about your products or services, and you start typing out the same explanation you’ve typed a hundred times before. Your process. Your pricing. Why you’re different. You try to make it feel personal, but honestly, you’re mostly copying from the last version you sent and tweaking a few details. Twenty minutes later, you hit send and move on to the next one.

Now imagine next week that same email only takes you three minutes. Not because you sent something generic, but because your AI assistant has begun to understand your business, your voice, and the specific things that matter to different types of prospects. The email that went out felt more personal than your copy-paste version ever did—because you actually had time to think about this particular person instead of grinding through the boilerplate.

And here’s where it gets interesting: the week after, your AI starts to flag things like, “This person mentioned they’re coming from a competitor—based on past conversations, you usually want to address the switching process upfront. Want me to add that?”

This is a promise of AI that many have made, and few have yet to unlock. To get this level of nuance out of your AI systems, it requires large amounts of context and examples for your AI to draw from, and the big P word you’ve probably heard before in this space - Prompt Engineering.

Maybe you’ve seen the courses. The YouTube tutorials. The LinkedIn posts about “mastering ChatGPT” that assume you have endless hours to become an AI expert on top of, you know, your actual job. Frankly, that can seem daunting.

Here’s the thing: you don’t need all that. There’s a simpler way.

Let the AI Figure Out the AI Stuff

That might seem counterintuitive, but stick with me.

The conventional wisdom says you need to master prompting techniques to get real value from AI. Learn the frameworks. Study the syntax. Practice crafting the perfect instructions.

Turns out, that’s not quite right. AI systems are actually better at optimizing and writing prompts than humans are—in some cases outperforming human-written prompts by up to 50% on complex tasks.

They understand their own quirks in ways we simply don’t. Or at least it would take PhD level study to get there. Does anyone here have an extra four years lying around to dedicate to deep academic study of probabilistic information models? I sure don’t.

So instead of spending your evenings learning how to talk to AI, it turns out you can just... ask the AI to get better at working with you. While you do your actual work.

This is learning by doing instead of learning then doing.

Compounding Productivity

Before we get into the how, let’s talk about why this approach - though it will take some up front time investment - is worth it, even when it feels like time is exactly what you don’t have.

You know how compound interest works with money: a small amount invested early grows exponentially because the returns generate their own returns. You can enact that same principle while working with AI systems like ChatGPT, Claude, Gemini.

Each time you finish a task with AI, you can capture what worked and roll that into the next interaction. You’re making a small deposit into a system that pays you back on every future interaction.

The first few rounds might feel like modest wins—saving 15 minutes here, catching an oversight there. Not exactly life-changing.

But those gains stack. By iteration six or seven, you’re not just working a little faster. You’re working at a completely different level.

The Approach: Iterative AI Prompting

Here’s the whole approach, step by step. Give it a shot this week, and let us know how it went.

Step 1: Pick one thing you do regularly.

Don’t try to transform your whole life with AI all at once. Pick one repeatable task—responding to inquiries, writing proposals, summarizing meetings, following up with leads, whatever.

The best tasks to focus on might fall into one or more of these categories

• High value workflow - Greater quality & efficiency in this area would have meaningful business upside

• Bucket drainers - busy work that has to get done but you don’t like doing it and you want to automate

• Network effects - many on your team can benefit from sharing this and iterating on it together

Ideally, you’re finding something that converges in these categories

Step 2: Do the work with AI prompts (as much as you can)

Open ChatGPT, Claude, or Gemini. Don’t overthink it. Upload your files, paste in previous examples, ask questions, go back and forth until you get something good. Tell it exactly what to change. Let it be messy. Messy is fine.

If it can’t quite get what you’re after, take over and finish the task manually, that’s part of the process.

Step 3: Ask the AI to capture what worked.

Here’s the key to the kingdom. When you’re done, ask this exact question:

“Based on my prompts in this chat, create a starter prompt I can use next time to get to this quality of output faster.”

That’s it. The AI will write detailed instructions based on everything it learned about your preferences, your context, and what “good” looks like to you.

Step 4: Save it somewhere you’ll actually use it.

You’ve got options here, from simple to sophisticated:

The quick and dirty way: Copy the prompt into a Google Doc or note. Next time, paste it into a new chat. It’s not elegant, but it works.

The better way: Save it as a reusable template in a custom GPT, a Claude Project, or a Gemini Gem. These let you start every new conversation with your custom instructions and examples already loaded—no copy-pasting required.

The best way - fit for purpose: Tools like Jasper, Missive (for email specifically), or Notion let you save custom prompts and templates for specific use cases like writing emails, managing marketing communications, and streamlining task completion, etc. These add cost, but if you’re doing high-volume work, they can be worth it.

Step 5: Keep feeding it your wins.

Upload your final deliverables as examples. Now the system doesn’t just know your instructions—it knows what good work actually looks like.

Rinse and repeat. Each cycle makes the next one faster. After three or four rounds, you’ll feel the difference.

What it Really Looks (and Feels) Like

All that sounds pretty straightforward, but let’s play it back through using that inquiry response example from the opener to fill out the picture further:

Round 1: Kind of a slog, honestly

Your first session takes a while. You paste in a few inquiry emails you’ve received recently and some responses you were happy with. You explain the different types of people who reach out—some are price-shopping, some are overwhelmed and need hand-holding, some are sophisticated and just want the facts.

You correct the AI’s tone three times. Too salesy. Too stiff. Weirdly formal. It’s a process.

But at the end, you ask:

“Based on my prompts in this chat, create a starter prompt I can use next time to get to this quality of output faster.”

You save what it gives you—whether that’s in a doc, a custom GPT, or wherever works for you. Done.

Rounds 2-3: Planting the seeds

Now you’re building momentum.

You open your saved prompt (or your custom GPT or Project), and the AI already knows your services, your tone, the different buyer types you see. You still tweak things here and there—adding a note to the instructions with specific phrasing to avoid, or clarifying how you handle pricing questions—but you’re getting to good output in half the time.

Most people stop here. And honestly, that’s a solid win already. But there’s more.

Round 4: Reaching the clouds

Something shifts around this point. You’re responding to inquiries way faster than before. But what’s more interesting is what that frees up your brain for (your organic intelligence, as it were).

You’re not having to grind through the mechanical stuff anymore—explaining your process for the hundredth time, remembering to include that paragraph about timelines, making sure you didn’t forget the call-to-action. That’s handled.

Instead, now you can spend your time thinking more about things you never had enough time for: What’s this person actually worried about? What’s between the lines in their message? What would get this conversation to moving towards a yes?

You’re starting to move from doing the work to directing the work.

Rounds 5-6 and beyond: Cooking in the castle (customer inquiries only please, no englishmen named Jack)

This is where it starts to get fun.

You update your instructions accordingly: “Before drafting, help me read between the lines of their inquiry. What might they really be asking? What objections should I address preemptively?”

And the system starts helping you apply that next level of thinking too. It notices patterns you hadn’t even articulated. It starts to flag things unprompted: “They mentioned they’ve been burned before—might be worth acknowledging that directly and explaining how your process is different.”

At this point, the AI isn’t just your typist. It’s more like a thought partner, helping you reach higher levels of consideration than you could. Hard to think critically when you’re just trying to grind through the inbox so you can shut down for the weekend.

Now that’s baked into the process, and the whole thing is happening faster than it did before.

Why This Works - Knowledge Compression

If you’re curious about why this works a bit under the hood, here’s the quick version.

Large language models are basically massive compression engines. They’ve taken an incomprehensible amount of human knowledge and squeezed it into something you can have a nice cordial chit-chat with.

When you follow this process, you’re adding your own layer of knowledge compression on top. Your communication style. Your expertise. Your organization’s weird quirks. Your stakeholders’ pet peeves. All of it gets condensed into a set of instructions and examples that constrain and direct the AI as it works.

And because each iteration adds to that foundation, it compounds. You’re not starting over every time. You’re building on top everything that came before.

The important thing to understand: this isn’t about replacing your judgment. It’s about encoding your judgment so the AI can extend it. You’re still the one who knows what good looks like. The AI just helps you get there faster.

The Bottom Line

You don’t need to take a course or watch a dozen YouTube tutorials to become an AI prompt Engineer. Turns out ChatGPT will always be better at it than you, and that’s okay. You just need to start doing your work with AI as your partner, and iteratively ask the system to get better at doing that exact thing as you go.

The people who are going to see major productivity gains in the AI era aren’t necessarily the ones who consumed the most educational content on LLMs. They’re the ones who started compounding sooner with small investments here and there.

The beanstalk starts with a single seed. Start planting yours this week!

This post was collaboratively brainstormed, researched, and drafted alongside AI (Claude Opus is our go-to lately for this sort of thing), then revised and edited heavily by real humans.

Conversint is a consulting firm helping SMBs and everyday humans unlock real value from AI and other modern tech. To learn more about our services and schedule a free initial consultation, visit our website at Conversint.ai.

If you learned something useful, consider sharing this post with someone else who may also benefit.

Share

For more insights and guidance like this delivered to your inbox regularly, consider subscribing.

Subscribe now