Most founders who’ve been at this for a few years have pieces.

Some strategic clarity. A decent presence. Content running, mostly. Owned channels being built. Customer work happening somewhere.

The pieces are disconnected. Nobody owns the full picture. Different parts run on different rhythms. Reporting covers what each piece did in isolation, not whether the whole thing is moving.

That’s an assemblage. Assemblages are fragile in a specific way.

What makes an assemblage fragile

The founder gets pulled into client work for a month and it frays. A key person leaves and part of the picture walks out with them. A new tool shows up, gets bolted onto the existing structure, and the whole thing gets more complicated without getting more effective.

I use a simple test for this: if you got hit by a bus tomorrow, could anyone in your business run the marketing for 6 months? If the answer is no, the system isn’t installed. The assemblage is being held together by you.

A Marketing Operating System is the opposite. Integrated, documented, connected, running on a rhythm the business can maintain with or without the founder’s constant attention.

The four components

Integration

Strategy, messaging, the engines, and the Hourglass diagnostic all connect to each other. Nothing sits in isolation.

When the strategy updates, the messaging updates with it. When the Hourglass surfaces a gap at Trust, the Brand Engine responds with specific work. When the Growth Engine tests a new offer, the Customer Engine updates onboarding.

In practice: one source of truth for strategy and messaging, engines that are explicitly defined and visibly connected to that strategy, and a shared vocabulary the whole team uses. When any of those are missing, changing one thing doesn’t change the things connected to it. The system drifts.

Cadence

Most small businesses have emergencies and campaigns. Cadence is different.

What it looks like in practice: a 30-minute weekly review covering what shipped, what moved, what’s blocking. A 60 to 90-minute monthly performance review against the 3 engines. A half-day quarterly planning cycle. A full-day annual strategy refresh.

Cadence is unglamorous. It’s also the most reliable predictor of whether a Marketing Operating System survives or decays over time.

Measurement

Five to seven metrics connected to business outcomes. Brand Engine: presence health, content engagement, list growth. Growth Engine: inbound volume by channel, conversion by channel, owned vs paid ratio. Customer Engine: repeat rate, referral rate, customer lifetime value.

Reported consistently, in context, against a goal. A report that tells a story: what happened, why it matters, what we’re doing about it. Not 16 numbers in a spreadsheet that nobody changes a decision based on.

AI as a leverage layer

This is where AI actually belongs, and most advice on this is either too enthusiastic or too dismissive to be useful.

AI can’t make strategic judgments about your market, your customer, or your business. It can’t produce your point of view. The thinking is still your job.

Where it does belong: research, production, reformatting, analysis, and the communication mechanics layer (follow-up, scheduling, first drafts). Installed on top of a working system, AI compounds advantage. Installed in the absence of one, AI amplifies the confusion that’s already there.

That distinction is the one most founders aren’t being given clearly right now.

What it looks like when it’s working

A professional services firm I worked with did the Founder Portrait work, rebuilt around a single service line, installed Strategy First, built out presence, content, owned channels, and a Customer Engine, then ran the full system for 2 years.

Revenue up 60% on lower marketing spend than before. Paid acquisition dependence cut significantly. Meaningful recurring revenue from the Customer Engine. And the founder can step away for 2 weeks without the system breaking.

Because it’s no longer being held together by his attention.

That’s a Marketing Operating System.

One thing to do this week

Do the bus test honestly. Write down everything about your marketing that only you know. Who the real ICP is, because the document is out of date. What the actual priorities are, because the quarterly plan never got finalized. Which conversations are in progress.

Whatever ends up on that page is the gap between the assemblage and the system. That page is the first draft of the Marketing Operating System document.

The Marketing Operating System is the final step of a seven-step framework I’ve been refining for over 20 years. The full system, from the Founder Portrait through the MOS, is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.

The Marketing Hourglass has 7 stages: Know, Like, Trust, Try, Buy, Repeat, Refer.

Most small businesses have systems for the first five. They know how to get found, how to build some trust, how to close. Then the marketing ends.

Repeat and Refer, the back half, get left to chance. Good work, happy customers, and hope.

That’s expensive. And it leaves most of the growth on the table.

What a customer is actually worth

A customer who buys, comes back, and refers is worth between 3 and 10 times a customer who buys once. That ratio shows up in the data of almost every small business that tracks it.

And yet. Acquisition gets the meetings. Acquisition gets the budget. Customer experience gets the leftovers.

I worked with a landscape services business at about $4 million in revenue. Growing through Google ads, word of mouth, and one partnership. The owner knew he had loyal customers but had never systematized any of the customer work. Within 12 months of installing a Customer Engine, it accounted for roughly 45% of total new revenue, up from about 10%. Paid acquisition spend dropped by a third. Because the back half of the Hourglass was finally doing its job.

Four things the Customer Engine does

Onboarding

The first 90 days after a customer buys is where the relationship gets established. Most businesses treat it as operations: deliver the thing that was sold, move on.

A structured onboarding process does something different. It confirms the customer made the right decision. It surfaces anything that needs fixing before it becomes a problem. And it creates the natural moment to ask for a review, a referral, or both.

Most businesses skip the ask entirely. The onboarding sequence is what makes it feel natural instead of awkward.

Repeat engagement

What specifically brings your customers back? Most businesses rely on the customer remembering to return. The Customer Engine removes that dependency.

Maintenance plans, seasonal offers, anniversary touchpoints, check-ins anchored to natural moments in the customer’s life. The landscape business introduced seasonal maintenance plans and converted about 40% of project customers. Recurring revenue went from essentially zero to a meaningful line.

That happened because they asked.

The referral system

Same 3 parts as the Growth Engine: a specific ask, at the right moment, with an easy path for the referrer. All 3 matter. Most businesses have none of them.

The right moment is right after something good, while the experience is still fresh. The landscape business built this properly. Referred customers went from about 10% of new work to 25% within 6 months. That’s a system, not luck.

Reactivation

A one-time outreach to every customer from the prior 3 years who hasn’t purchased anything new. Simple, direct, personal note from the founder.

The landscape business converted about 8% of that list into some form of re-engagement within 90 days.

Reactivation is probably the highest ROI marketing move available to most small businesses. Almost nobody does it, mostly because it feels like admitting you lost touch. Reframe it: it’s a welcome reconnection, and customers respond to it that way.

What the Customer Engine actually powers

This is the part most founders miss. The Customer Engine doesn’t just produce direct revenue from existing customers. It feeds every other engine you have.

The Trust stage needs customer stories. The Customer Engine produces them. The Refer stage needs actual referring behavior. The Customer Engine systematizes it. The content engine needs real situations and wins. The Customer Engine surfaces them.

Under-investing in the Customer Engine under-powers everything else. Fixing it lifts the whole system, not just retention.

One thing to do this week

Write your referral system in one sentence.

If it turns into a paragraph of caveats, or “we don’t really have one,” that’s your answer. And it tells you exactly where to start.

The Customer Engine is step 6 of a seven-step system I’ve been refining for over 20 years. The full framework is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.

If your largest paid channel disappeared tomorrow, platform shuts down, algorithm changes, cost doubles, your pipeline is gone inside 30 days.

If that’s true, you don’t have a Growth Engine. You have a rented pipeline.

This is the situation most founders are in. Paid ads on two or three platforms. Paid social. Maybe a paid directory. When the credit card stops, the leads stop. The business has revenue but no predictability. It has a dependency.

Owned vs rented

An owned channel is one you control. You decide who’s on it, what reaches them, when. No platform change can touch it.

A rented channel is one someone else controls. You pay for access. You play by their rules. When the rules change or the price goes up, you adjust or you disappear.

The difference compounds over time. A business that builds owned channels for 5 years has compounding value. A business that rents for 5 years has 5 years of expenses. Same spend, completely different position.

The four owned channels

Email

Email has been declared dead roughly once a year for 15 years and keeps working anyway.

A founder who builds a qualified email list over 5 years has direct, reliable, owned access to their audience at zero marginal cost per send. No paid channel comes close to that math.

The list has to be qualified, built from people who asked to be on it. It has to be used consistently. And it has to be treated as a content surface, not a sales channel. The same principles that make content work apply here: genuine point of view, useful, specific.

Most businesses underuse email because it feels unfashionable. That unfashionability is the tell. The channels that feel unfashionable and still work are the ones smart operators quietly compound in.

Referral systems

Referred prospects arrive pre-trusted. They close faster, they’re less price sensitive, and they’re more likely to refer others. Most small businesses have no referral system. They have referrals that happen accidentally.

A real referral system has 3 parts: a specific ask, made at a specific moment, with a specific easy path for the referrer to take. All 3 are necessary. Most businesses are missing at least 2.

The ask needs to be made. Customers don’t refer unless asked because it’s not obvious to them that you want referrals. The moment matters: right after a customer experiences something good is when the ask lands. And the path needs to be easy enough that referring requires almost no effort.

Partnerships

Non-competing businesses that serve the same ideal client are the most underused lead source in small business marketing.

An accounting firm’s ideal client also needs a business lawyer, a financial planner, a banker, an insurance broker. Each of those providers has a list of the same customers. Two or three real partnerships beat 20 casual ones.

A structured partnership has named partners, defined criteria, a regular rhythm of contact, and a way to track what’s being exchanged. Partnerships are work. They compound once they’re real.

Direct relationships

Networking, speaking, association involvement, in-person participation. The oldest channel in the book, and it still produces the highest-intent leads in most categories.

A prospect who hears the founder speak at an industry event arrives at the buying conversation miles ahead of where a paid lead arrives. The trust is largely pre-built.

Direct relationships don’t scale the way email or content scale. They scale with founder effort. Founders who invest in them consistently find that the Growth Engine runs mostly on relationships 2 years in.

Where paid actually belongs

Paid works when it amplifies something already working. If the content is converting organically, paid can extend its reach. If the messaging is landing, paid can get it in front of people it otherwise wouldn’t reach.

Without those foundations, paid produces expensive activity that doesn’t convert. Every founder has seen that at least once.

The healthy ratio for most small businesses: roughly two-thirds of new customer flow from owned channels, one-third from paid amplification. A business running the opposite ratio is fragile, even if the current economics look fine.

One thing to do this week

List every lead source that produced revenue in the last 12 months. Mark each one owned or rented. Count the ratio.

If rented is more than half, the Growth Engine is the priority. Start with the owned channel closest to working but undeveloped. That’s usually email or referral.

The Growth Engine is step 5 of a seven-step system I’ve been refining for over 20 years. The full framework is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.

The right move was the opposite: publish less, go deeper, commit to 3 content pillars.

I see this pattern constantly. Founders who aren’t getting results from content assume the problem is volume. So they add more posts, more channels, more tools. And they get the same results, faster.

Producing more generic content doesn’t fix a content problem. It amplifies it.

The actual problem

Most small business content doesn’t have a job. It’s a series of posts with no spine underneath. Topics that seemed interesting that week. Updates that felt like they should be covered. Technically useful stuff that adds up to nothing.

In a market where AI is generating generic content at industrial scale, being part of the noise layer is bad for your brand. The customers worth winning have started to recognize it and tune out.

Content that actually works does one thing: it earns trust before the customer has to talk to you. It signals that you understand their situation, you’ve thought about it seriously, and you have something specific to say.

Pillars, not posts

Pick 3 content pillars anchored to your ideal client’s real problems. Every piece of content you publish goes to one of them.

I know how this sounds. Organization. A content calendar thing. It’s actually the hardest strategic decision most founders avoid making.

Most businesses publish what the founder was thinking about that week. After a few years you have a body of work with no accumulated weight. A prospect can’t tell what you’re actually expert in.

Three pillars held over 2 or 3 years produces a different result. The body of work has shape. The depth on each pillar becomes visible, and that visibility is what earns trust.

Three is the right number. Two is too narrow. Four dilutes. Three works.

Each pillar has to pass 3 tests: anchored to a real customer problem, an area where you have genuine depth, and one you can publish against for 3 years without getting bored. If it won’t survive that last test, it’s a topic, not a pillar.

Hubs, not archives

Content organized under hub pages compounds over time. Content organized as a reverse-chronological blog buries your best work within weeks.

The reverse-chronological blog is an artifact from when blogs were journals. It made sense then. When content is meant to be a long-term asset serving both readers and AI retrieval systems, it doesn’t.

Under hub pages, your best work stays discoverable and accumulates authority. When you publish something new, link it to the appropriate hub and update the hub to reference it. Over time the hub becomes a genuine knowledge center. The blog archive becomes a graveyard.

Repurposing, not more production

The founders who win on content get maximum leverage out of each substantial piece. Volume isn’t the advantage.

The model: one substantial piece per week or two, repurposed into 8 to 10 smaller assets. A podcast episode becomes a hub page article, a few LinkedIn posts, one email to the list, a short video. A long article becomes an email series, a handful of social posts, eventually a book chapter.

This is where AI actually earns its keep. Taking original thinking and adapting it across formats is something AI does well. Producing original thinking from scratch isn’t. Keep the thinking yours. Use AI for the reformatting.

The point of view problem

The market is full of AI-produced content that reads like AI-produced content. Generic, balanced, readable, forgettable.

The content that still earns attention, gets remembered, and gets shared has a point of view. It takes a position. It says something the customer hasn’t heard, or says something familiar in a way that makes it land differently.

AI can’t produce a real point of view because it’s averaging the existing corpus. Your specific perspective isn’t in there.

Use AI to produce. The thinking is still your job.

Content without a point of view was dismissible in 2020. It’s invisible in 2026.

One thing to do this week

Name your 3 content pillars on one page. If you can’t narrow to 3, the narrowing is the work. Three is not a formatting choice. It’s the strategic constraint that forces real decisions.

Content strategy is step 4 of a seven-step system I’ve been refining for over 20 years. The full framework is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.

Does your business show up?

I’ve run this test with dozens of small business owners in the last year. Most of them disappear completely. Some show up but get described in ways that would make a prospect walk the other direction. A handful get it right.

The ones who get it right aren’t doing anything exotic. They’ve just built a presence that works the way presence has to work now, which is different from how it worked five years ago.

Presence used to have one job

For the first 20 years of the commercial web, presence meant one thing: Google could find you. Get the SEO right, show up in search, done.

That’s still necessary. It’s just not sufficient anymore.

A working presence in 2026 has to pass three tests, and most small businesses are failing at least one of them without realizing it.

Job 1: Findable

Can the right customer, searching for the right thing, actually find you? The mechanics have shifted. Less about keywords stuffed into pages, more about genuine topical authority built over time. But the test is the same.

Here’s the part most people miss: findable now means findable in three places. Traditional search (Google, Bing). Social search (people searching inside platforms). And AI-mediated search, ChatGPT, Perplexity, Google’s AI Overviews, and the vertical AI tools your customers are quietly starting to use for research. Each one pulls from different signals. Build for only one and you’ve got gaps.

Job 2: Credible

When a prospect lands on your site, does the site do its job? Does it speak to their situation in their language? Does it show real proof that you’ve done this work for people like them?

I see beautiful websites every week that fail this test completely. Design isn’t the problem. Most of them look great. The problem is there’s nothing there. Generic copy, stock photos, and a contact form. A plain site with deep, specific proof of real work outperforms a polished site with nothing behind it every time.

Job 3: Retrievable

This is the new one, and it’s the one catching businesses off guard.

When an AI assistant answers a question your customer asks, “who should I hire to do X in Y city” or “what should I look for in a contractor for Z,” does your business come up? And when it does, is the description accurate?

AI systems build their answers from whatever you’ve put out publicly. Thin website. Generic content. Missing structured data. Weak third-party presence. The AI either won’t find you or won’t know how to describe you. Being un-retrievable is just the new version of being un-findable. The customer moves on and you never know it happened.

Three things to fix first

Your website

Most small business websites are expensive brochures. They describe the business but don’t sell it. Four things fix most of them: a clear core message above the fold, the ideal client named in their own language, specific proof material, and one obvious next step. Not “contact us.” One low-friction action for the person who’s ready to move.

Hub pages

A hub page is a deep, authoritative page built around one specific topic: a core service, a core customer problem, a category you want to own. Not a blog post. A real resource that earns its place as the best answer on that topic.

Search engines rank them. AI systems cite them. And they give your content something to cluster around instead of floating independently. If your site doesn’t have hub pages, you’re competing on a level playing field with everyone else in your category. Hub pages tilt that field.

Your presence beyond the site

AI doesn’t build its picture of your business from your website alone. It pulls from your Google Business Profile, industry directories, third-party reviews, and mentions across the web. Most small businesses treat this as low-priority busywork. It’s actually the scaffolding holding everything together.

A business with a solid website and strong third-party presence will beat a business with a great website and weak external presence in AI-generated answers. Every time.

Do the test today

Open an AI assistant. Type three questions your ideal customer might ask before hiring someone in your category. Screenshot what comes back.

That’s your baseline. That’s what your prospects are seeing right now. It tells you exactly where to start.

Online presence is one of the seven steps in the framework I’ve been refining for over 20 years. The full system is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.

This is the most common condition in small business marketing. And it’s almost never caused by lack of effort or thin budgets. It’s caused by the absence of a strategic foundation the tactics can actually build on.

What founders mistake for strategy

Most founders with a tactics problem think they have a strategy. They almost never do.

What they have is a list of tactics they’re running, opinions about each one, and a history of what did and didn’t work. That’s not a strategy. A strategy is a coherent answer to three questions:

Who exactly are we for? What do we do that the alternatives don’t? What’s the one sentence that ties those two things together?

Without those answers, the tactics underneath can’t compound. They just take turns failing.

Strategy First: the three pieces

The strategic foundation has three parts. All three have to exist. Any one of them alone isn’t enough.

The ideal client

A persona isn’t an ideal client. A demographic isn’t an ideal client. “Small business owners between 35 and 55 who value quality” is a description, not a strategy.

An ideal client is a specific type of customer, in a specific situation, whose problem you’re uniquely positioned to solve better than the real alternatives they’re actually considering.

Here’s what specificity looks like in practice: a home services company whose ideal client is “owners of 20-plus-year-old homes in zip codes where houses sell for over $800,000, who’ve lived there more than 3 years and are thinking about aging in place.” That’s a strategy. Every downstream decision, where they advertise, what their photos show, how they price, what they stop offering, can align to that specific person.

The riches are in the niches. That was true when I wrote the original Duct Tape Marketing. It’s more true now. In a market where AI makes it trivially easy to produce generic content for generic audiences, the only marketing that gets through is the marketing clearly made for someone specific.

Differentiation

Two mistakes come up constantly. Claiming differentiation that isn’t actually different (quality, service, experience: every business claims these). And describing differentiation against the wrong competitor.

Your customer is rarely choosing between you and the obvious direct competitor. They’re choosing between you and doing nothing, a different category of solution, or doing it themselves. Your differentiation has to land against that actual set of alternatives.

Differentiation is also a commitment. If you claim to be the firm that does the deepest strategic work before any execution, you can’t also take an emergency project on Monday and deliver by Friday. The claim requires you to turn down certain work. That’s the real test: does your differentiation require you to say no to something?

The core message

One sentence. In the customer’s language. Describes who you’re for and why they’re in the right place.

It has to pass 3 tests. Clear (a smart 12-year-old should understand who you serve and what you do). Different (it can’t be lifted and pasted onto a competitor’s site without anyone noticing). Credible (the customer believes it).

Clever is a tagline. The core message is clear. They can be the same thing. They usually aren’t.

The Marketing Hourglass

Strategy First also gives you the diagnostic lens you’ll use for everything that comes next: the Marketing Hourglass.

Most people were taught to think about the customer journey as a funnel. Leads in the top, customers out the bottom. It’s useful for a narrow slice of the work and dangerously incomplete for the whole picture.

Real growth for small businesses happens inside an hourglass, because the most valuable customer activity happens after the sale. The 7 stages: Know, Like, Trust, Try, Buy, Repeat, Refer. The hourglass widens again after Buy. That’s the part most small businesses ignore, and it’s where the highest-value growth actually lives.

The diagnostic is simple: find the stage where things are leaking and fix it before you build anything new on top.

One thing to do this week

Write your core message. One sentence. Customer’s language. Run it through the 3 tests: clear, different, credible.

If it can’t pass all three, that’s the strategy work. Everything else waits until it does.

This is step two of a seven-step system I’ve been refining for over 20 years. The full framework is in my new ebook, “7 Steps to Small Business Marketing Success.” Get it at dtm.world/7steps.