Speed It Up a Little: A Content Marketing System for Small Business Owners

Law Smith

Founder & President, Tocobaga

Law Smith is the Founder and President of Tocobaga, a Tampa-based ROI-focused marketing agency and SMB advisory. He has strategically advised 1,000+ small and medium businesses, executed 600+ integrated marketing campaigns, and written 30+ business plans across industries ranging from professional services to e-commerce to B2B technology. Tocobaga is a Google Partner and Squarespace Gold Partner headquartered in Ybor City.

The belt starts slow. On the second-season premiere of I Love Lucy, broadcast on September 15, 1952, Lucy Ricardo and Ethel Mertz stand at a wrapping station inside a candy factory, white caps on, sleeves rolled, a conveyor belt delivering chocolates one at a time.* The forewoman gives them the only instruction that matters: one unwrapped piece reaching the packing room means they are fired. Then she calls for the belt to roll.

For about a minute, the job is easy. Pick up a chocolate, wrap it, set it back down. Then the belt speeds up. The chocolates arrive in pairs, then in clumps. Lucy starts eating them. Then both of them start hiding them. When the belt stops and the forewoman returns to inspect a spotless, empty station, she tells the two of them they are doing splendidly and orders the belt sped up a little. By the end of the scene they are stuffing chocolates into their caps and down the fronts of their uniforms, cheeks packed, eyes wide, while the belt keeps coming.†

TV Guide put that scene second on its 1996 list of the 100 most memorable moments in television history, and a 2013 Paley Center special ranked it first among its 30 funniest moments in television history.† It is funny because everyone has been on that belt. Not in a candy factory. In the part of the week where a business owner is supposed to be doing content.

The belt is the calendar. The chocolates are the posts, the newsletters, the reels, the LinkedIn updates somebody said had to go out daily. The forewoman is the algorithm, or the consultant who said consistency is everything, or the competitor across town who seems to publish constantly. And the owner, who has an actual business to run, a payroll to make and a client waiting on a callback, is eating the chocolate. The posts get shorter and thinner. Then they stop. Then there is a four-month gap on the company blog that every prospect can see.

Tocobaga, the lead generation agency and SMB advisory Law Smith runs out of Ybor City in Tampa, publishes on SolvingHow.com, and across 1,000+ advisory conversations with owners the chocolate factory confession comes up more than almost any other topic. The owner knows content works. The owner has watched a competitor's article outrank the firm for a question the firm answers better in person every week. The owner tried, for a while, and then the belt won.

A generation before Lucy's belt, in a plant in Highland Park, Michigan, a different conveyor solved the same problem in the opposite direction. That belt did not speed up on the worker. It brought the work to the worker, in a fixed order, one task per station, and it cut the time to build a car from a full working day to about an hour and a half. The difference between those two belts is the difference between content marketing that eats an owner's week and a content marketing system for small business that runs on two hours, on the same afternoon, every week. Building the second belt is a design problem, not a stamina problem.

Why does content marketing feel impossible for a small business owner?

Because most owners run content the way Lucy ran the wrapping station: by hand, one piece at a time, with the belt setting the pace. Content feels impossible when the work has no stations, so every piece starts from zero and every interruption resets the clock.

The numbers on how long a single piece takes are not kind. Orbit Media has surveyed bloggers every year since 2014, and its 2026 edition, drawn from 1,042 content marketers, puts the average time to write one post at 3 hours and 20 minutes, for an average post of 1,312 words.‡ For a professional writer with a clear desk, that is an afternoon. For the managing partner of a two-partner family law firm in Tampa, it is an afternoon that does not exist, which is why the drafting happens at 9:40 on a Tuesday night, in 25-minute fragments between emails.

Fragments are the expensive part. Gloria Mark, a professor of informatics at the University of California, Irvine, had observers shadow information workers for days at a time and found it took an average of 23 minutes and 15 seconds to get back to a task after an interruption.§ Three interruptions during a 3-hour-and-20-minute writing session is more than an hour lost to re-finding the thread. The owner experiences that as being bad at writing. It is not a talent problem. It is a station problem: a wrapping station with no walls, set in the middle of the factory floor where anyone can walk up and hand over another chocolate.

The payoff for all that effort has been shrinking. Only 13.9 percent of the bloggers Orbit surveyed in 2026 reported strong results from their blogging, which the survey calls an all-time low, even as 92.4 percent of them now use AI somewhere in the process.‡ More tools, more output, fewer results. The Content Marketing Institute's 2026 B2B survey of 1,015 marketers tells the same story from the other side of the desk: 39 percent name resource constraints (time, people, budget) as a top challenge, 40 percent say their hardest problem is creating content that prompts a desired action, and only 12 percent rate their content marketing as highly effective, meaning it exceeded its goals.‖ Those are marketing departments with headcount. A PE-backed HVAC platform with 11 locations and one regional marketing coordinator is working with a fraction of that, and a solo management consultant is working with none of it.

So the owner is not imagining the difficulty. The belt really is faster than one person wrapping by hand. The mistake is assuming the answer is a faster person.

What is a content system and how is it different from a content calendar?

A content calendar is a list of dates with topics written on them. A content system is a set of stations: where ideas come from, who drafts, who edits, where each finished piece travels next, and how the business finds out whether any of it produced a client. The calendar says when the belt runs. The system decides what rides on it and who touches it.

The cleanest picture of the difference comes from a slaughterhouse. A Ford Motor Company foreman named William Klann, known on the floor as Pa Klann, visited the Swift and Company meatpacking plant in Chicago and watched carcasses move past stationary butchers, each man making the same cut on every animal that passed, nobody walking anywhere. Swift called it a disassembly line. What caught Klann's eye was one worker removing the same piece over and over without ever leaving his spot, and he went back to Detroit with the idea of running that process in reverse.¶

Ford tried it first on the smallest thing it built. Around April 1, 1913, the company put flywheel magnetos on a moving line at its Highland Park plant. Before the change, one worker built one magneto from start to finish in roughly 20 minutes. On the moving line, with the job split among workers who each did one step, a finished magneto came off every 13 minutes, and after some tuning, every five.** By late summer the method had spread to engines, transmissions and chassis, and October 7, 1913, is the date usually given for the first moving automobile assembly line, at Highland Park.†† By early 1914 the separate lines had merged into three continuous ones, and a finished Model T came off the end every 93 minutes, against 12 and a half hours per car under the old method, where the car sat still and the workers walked to it.**

The part owners tend to miss is what Ford did not do. Ford did not hire faster workers. Ford did not tell the existing workers to speed it up a little. The company moved the work to the worker, broke the job into 45 steps, fixed the order, and let the belt carry the car past people who no longer had to think about what came next.¶ The price of a Model T fell from $700 in 1910 to $350 in 1917, and on January 5, 1914, the company raised its daily wage from $2.34 to $5 and cut the shift from nine hours to eight, because the line had made each hour worth more.††

That is the whole argument for a content system in one factory. Lucy's belt sped up on the worker. Ford's belt brought the right part, in the right order, to a worker who did one thing well. An owner whose customer questions arrive in a ledger, whose first draft comes off an AI station in 20 minutes, whose edit is a fixed checklist, and whose finished piece becomes a newsletter and five short posts by default is running Highland Park.

The comparison below sets Lucy's belt next to Ford's belt, station by station.

Lucy's belt versus Ford's belt: the hand-wrapped content habit next to the station-based content system.

Station	Lucy's belt (the hand-wrapped habit)	Ford's belt (the content system)
Channels	Be everywhere: blog, LinkedIn, Instagram, YouTube, a podcast, all fed by the owner	One home channel (the website's answer pages) and one delivery channel, for six months minimum
Topics	Brainstormed on a Tuesday night from a blank page	Pulled from a question ledger logged within two minutes of every sales call
Time	Fragments between emails, each restart costing about 23 minutes	One protected two-hour block, same afternoon every week, phone in a drawer
Drafting	Owner types everything from scratch, or AI publishes unedited	Owner dictates, AI cleans the transcript into a draft, owner runs a fixed edit checklist
What the machine never writes	Nothing is off limits, so invented figures and citations slip through	Prices, fees, legal claims, guarantees, statistics and citations are human-only
Cadence	Daily posts for three weeks, then a four-month gap	One long piece a month, a ladder every week: weekly presence, monthly depth
Repurposing	Each post is a one-off; nothing is reused	Every long piece becomes a newsletter, five short posts and a 30-second video by default
Measurement	Likes, impressions and follower counts	Visits, leads, signed clients and revenue per piece, from a required source field and the landing page

Which content channel should a small business focus on first?

One channel, chosen by where the next client already looks, and the channel is almost always the business's own website. Everything else is a distribution stop for work that lives at home.

The logic is Ford's, not Lucy's. Highland Park did not run five lines building five different cars. It ran one line building one car and then let dealers in a thousand towns sell it. A business that tries to feed a blog, a LinkedIn page, an Instagram grid, a YouTube channel and a podcast at the same time is running five belts with one worker, and the worker is the owner. The Content Marketing Institute found that 39 percent of B2B marketers name time, people and budget as a top constraint even inside organizations with dedicated staff.‖ The fix for a constraint is not more belts.

The website wins the first slot for an unglamorous reason: it is the only channel the business owns outright, and it is the only one that answers a question at 2 a.m. for a stranger who has never heard of the firm. A LinkedIn post disappears from the feed in a day. An answer page on a law firm's site that explains what a contested divorce in Hillsborough County actually costs will still be pulling intake forms in three years, and it will be what Google and the AI engines quote when someone asks the question out loud. Every email campaign worth sending points somewhere, and the somewhere should be a page the business controls.

For a two-partner law firm, that means answer pages and a monthly email to past clients and referral sources. For a mid-sized manufacturer, it means application and specification pages (what the part does, what it costs to tool, what lead times look like) plus a LinkedIn presence for the owner, because that is where plant managers and procurement people spend their scroll. For a solo management consultant, it means one long article a month on the site and a personal LinkedIn account that reposts pieces of it, because the consultant's face is the brand. Social media has its own set of problems and its own post on SolvingHow.com; here it is the delivery truck, not the factory.

The rule Tocobaga uses with owners is simple: one home channel, one delivery channel, and nothing else until the first two have run for six months without a missed week. A PE-backed HVAC platform with 11 branches and a single marketing coordinator does not get an exception. It gets a stricter version, because 11 Google Business Profiles and 11 location pages are already more belts than one coordinator can wrap.

Where do content ideas come from when there is no time to brainstorm?

From the questions clients already ask, written down at the moment they ask them. The content list for the next 12 months is already sitting in the owner's sent folder, the intake notes and the last 30 sales calls; nobody needs a brainstorm, they need a ledger.

Klann's insight in that Chicago slaughterhouse was that the material should come to the station, not the other way around.¶ Topics work the same way. The owner who sits down on a Tuesday night to think of something to write about is walking to the car. The owner who keeps a running note on the phone, one line per question, logged in the two minutes after a call ends, has the car coming down the line. "What happens to the house if we file first?" "Can the two of us use one lawyer?" "Why does the quote go up 30 percent if the order drops below 500 units?" "What does a fractional engagement cost, and what do I actually get for it?" Each one is a page. Each one came from a real person with money, which is the only kind of keyword research that is free.

The reason this matters is that most content fails before it is written, at the topic. Ahrefs studied roughly 14 billion pages in its index in 2023 and found that 96.55 percent of them get no traffic from Google at all, and the first cause the study names is that nobody was searching for the topic in the first place.‡‡ Every one of those pages took somebody 3 hours and 20 minutes. A question ledger cannot produce a page nobody wants, because the question has already been asked by a paying customer.

The ledger also lines up with where search is going. Semrush tracked more than 10 million keywords through 2025 and found that the queries which trigger Google's AI Overviews tend to be longer and more specific, and that nearly 60 percent of them draw 100 or fewer searches a month.§§ Those are not the fat keywords an agency pitch deck shows. They are the exact, awkward, specific questions a client asks on a call, which is to say they are the ledger. A regional 3PL that writes one plain page answering "what does it cost to store 40 pallets a month in Tampa" is not competing with the logistics giants for "warehousing"; it is answering the question the giants do not bother with, for the buyer who is about to call someone.

The practical version: a shared note titled Questions, visible to everyone who talks to customers. One line per question, verbatim where possible, with the date. The first move is to go back 30 days through sent email and call notes to seed it on day one, and that first pass usually produces 40 to 60 questions. At one strong piece a month, that is four years of topics gathered in an hour. The belt never runs empty again.

Does batching content actually save time?

Yes, and the saving is larger than the arithmetic suggests, because batching removes the restart cost rather than the work. Writing three pieces in one sitting is not three times as fast as writing one; it is faster than that, because the owner pays the 23-minute re-entry toll once instead of three times.§

Ford's magneto line is the cleanest demonstration ever run. The same workers, the same parts and the same factory went from one finished magneto every 20 minutes to one every five, and the only variable that changed was that each worker stopped switching tasks.** Nobody got faster. The switching disappeared. Gloria Mark's interruption research is the modern version of the same finding: the cost of knowledge work is not the work, it is the resumption.§ An owner who drafts in fragments across a week is paying the resumption cost on every fragment.

Batching also makes the stations visible, and that is where the real design work happens. A content month with stations looks like this. Station one, the first 15 minutes of the first block: pull the top three questions off the ledger and pick one. Station two, the rest of that block: dictate the answer into a phone the way it would be explained to a client across the desk, then have an AI tool clean the transcript into a first draft with question headings. Station three, the second block: the human edit, where the owner adds the specific number, the war story, the price range, the one sentence a competitor would never write. Station four, the third block: the repurposing pass, where the finished piece is cut into a newsletter, five short posts and a video script, and all of it is scheduled. Station five, the fourth block: the measurement sheet, and the ledger refilled for next month. Four blocks, in that order, every month.

The order is the point. Lucy's station failed because every chocolate was an independent decision. Ford's line worked because no worker ever had to decide what came next. An owner who sits down to "do content" with no stations is deciding what to do 40 times in two hours. An owner with the stations fixed makes one decision, at station one, and the belt carries the rest.

Can AI write a small business blog post?

AI can write a competent first draft in minutes, and it cannot write the part that gets a business hired. The division of labor is the whole system: the machine runs the draft station, the owner runs the edit station, and the pieces that must be true never touch the machine.

The best evidence on what AI actually does to writing work comes from two MIT economists, Shakked Noy and Whitney Zhang, who ran a controlled experiment on 453 college-educated professionals (marketers, grant writers, consultants, HR managers) doing realistic writing tasks. Half got access to ChatGPT. Time to complete the task fell 40 percent, quality as rated by independent evaluators rose 18 percent, and the gap between the weakest writers and the strongest writers narrowed, because the weakest writers gained the most.‖‖ For an owner who hates writing, that last finding is the one that matters: the tool helps most exactly where the owner is weakest. The study's authors also warned that real-world gains would be smaller once fact-checking and prompt-writing time were counted, which is the honest caveat.‖‖

Here is where the caveat becomes a rule. Google's own documentation on helpful content asks whether a page clearly demonstrates first-hand expertise and a depth of knowledge, whether it offers information beyond the obvious, and whether it provides substantial value compared with other pages in the results; it also lists, as a warning sign, producing lots of content on many topics in hopes that some of it performs.¶¶ A model trained on the public internet can produce the obvious. It cannot produce what happened in the owner's conference room last Thursday, because it was not there. So the human edit station has a fixed checklist: add one specific client scene (anonymized), add one number from the business's own books, add the price range a competitor would hide, strip every sentence that could appear on any other firm's website, and read the whole thing aloud once.

Then there is the list of things the machine never writes, and it is not negotiable: prices, fees, legal and regulatory claims, guarantees, statistics, and citations. A machine will invent a case name or a percentage with total confidence, and a law firm that publishes an invented citation has a bar complaint waiting, not a marketing problem. The machine drafts. The owner decides what is true. AI copy that sounds like a robot has its own post on SolvingHow.com; the short version is that specificity is the only cure, and specificity only comes from the human station.

What is a content repurposing ladder?

A repurposing ladder is the fixed set of smaller pieces that every long piece is cut into, in the same order, every time. One long post becomes one newsletter, five short social posts and one 30-second video script, and the owner never decides whether to make them because the ladder is the default.

This is the step where Ford turned three separate lines into one continuous line in early 1914 and the output went from a car every 12 and a half hours to a car every 93 minutes.** The magneto line alone was a nice improvement. The merge was the revolution. In content, the long article is the magneto. The merge is the ladder, and the ladder is where a two-hour week starts producing what looks, from the outside, like a full-time content operation.

A worked ladder for a 1,800-word answer page titled "What does a contested divorce actually cost in Hillsborough County?" runs like this. The newsletter is the first 250 words of the page plus one paragraph of context the page does not include (a recent change in filing fees, say) and a link. Social post one is the single most surprising number on the page, stated plainly. Post two is the question in the heading, with a one-line answer. Post three is the client scene, anonymized, in three sentences. Post four is the one mistake the page warns against. Post five is the rule of thumb the page ends with. The video script is the owner, hips up, 45 degrees off camera, answering the heading question in 75 words to a producer off screen. Every one of those pieces points back to the page, which means every one of them is working for the home channel, not for the delivery channel.

The Edelman and LinkedIn thought leadership study of 3,500 B2B decision makers explains why the ladder is worth the 20 minutes. Seventy-three percent of buyers said thought leadership is a more trustworthy basis for judging a firm than its marketing materials, 75 percent said a piece of thought leadership led them to research a product or service they had not been considering, and 60 percent said they would pay a premium to work with a firm whose thinking they trusted.*** The same study found only 15 percent rated the thought leadership they see as very good or excellent.*** The ladder does not make a piece better. It puts the one good piece in front of the same buyer five more times, in the formats the buyer actually reads, at a marginal cost of almost nothing.

A manufacturer's ladder looks different in content and identical in structure: a page on why a part's price changes with order quantity, a newsletter to the 400 procurement contacts in the CRM, five LinkedIn posts from the owner with one number each, and a 30-second video shot on the plant floor next to the machine. One afternoon, six weeks of presence.

How much time should a small business owner spend on content each week?

Two hours, in one block, on the same afternoon every week, protected the way a court date or a client meeting is protected. Less than that and the system never gets its stations built; more than that and the owner is subsidizing marketing with the hours that run the business.

The two-hour figure is not arbitrary. Orbit Media's 2026 survey puts the average post at 3 hours and 20 minutes of writing, and its earlier editions found that bloggers who spent six or more hours on 2,000-word pieces roughly doubled their odds of reporting strong results.‡ Read quickly, that looks like an argument for more time. Read carefully, it is an argument for a system: the two-hour block does not produce a finished long piece every week. It produces one long piece a month, assembled across four blocks, with the other three blocks feeding the ladder, the ledger and the measurement sheet. Four two-hour blocks is eight hours a month on one piece plus its ladder, which is above the six-hour threshold the survey associates with strong results, and it fits inside the week of a person who also has to make payroll.

The block has to be defended, and the defense is physical. Gloria Mark's 23-minute resumption cost means a two-hour block with four interruptions is a one-hour block.§ So the block goes on the calendar as a client meeting. The phone goes in a drawer. For a two-partner firm, one partner takes the block and the other covers the phones, and they alternate months. For a PE-backed platform with a marketing coordinator, the coordinator owns the stations and the general manager of each branch owes the ledger ten questions a month, in writing, as a standing agenda item on the operations call. For a consultant working alone, the block is Friday from 1 to 3, and the autoresponder says so.

What the block is not: a license to publish daily. Orbit's long-run data shows the share of bloggers publishing two to six times a week fell from 28 percent in 2014 to 12 percent by 2022, while monthly publishing roughly doubled, and the survey lists publishing monthly or less among the practices tied to weaker results.‡ Those two facts sound contradictory until the system is in view. The right cadence for an owner is one strong long piece a month plus the ladder, which puts something in front of the buyer every week without requiring a long piece every week. Weekly presence, monthly depth. Lucy's forewoman wanted more chocolates per minute. Highland Park wanted more finished cars per day. Those are different goals, and only one of them pays.

Is it better to publish more content or better content?

Better, and the data is not close. A business that publishes one specific, first-hand, well-edited piece a month will outperform one that publishes four generic ones, on search, on AI citations and on the only metric that pays the rent, which is a buyer deciding to call.

The volume strategy has already been run at scale, and the results are in: 96.55 percent of 14 billion pages with no Google traffic, and 13.9 percent of bloggers reporting strong results in a year when 92.4 percent of them used AI to produce more.‡‡‡ Output went up. Results went down. The belt sped up, and the chocolates ended up in everyone's hats.

Google has said as plainly as a search engine ever says anything what it is looking for instead. Its helpful content guidance asks whether a page provides original information, reporting, research or analysis; whether it offers insightful analysis beyond the obvious; whether it clearly demonstrates first-hand expertise and depth of knowledge; and whether it adds substantial value compared with the other pages already ranking.¶¶ The same document lists extensive automation to produce content on many topics as a warning sign.¶¶ Every question on that list is a question about specificity, and specificity is cheap for an owner and impossible for a content mill. The two-partner firm knows what a contested divorce costs in Hillsborough County because it filed three last month. The content mill is guessing from a national average.

Buyers have the same preference. In the Edelman and LinkedIn study of 3,500 B2B decision makers, 73 percent said a firm's published thinking is a more trustworthy basis for judging it than its marketing materials, and only 15 percent rated the thought leadership they actually see as very good or excellent.*** That gap is the opening. A procurement office or a general counsel's inbox can tell when a piece was written by someone who has done the thing, and it can tell when it was not.

Tocobaga's rule across 600+ campaigns is that one piece with a real number, a real scene and a real price range beats ten without them, and the measurement sheet backs it up every quarter. The quality bar is not literary. It is a checklist with five lines: a question in the heading, a direct answer in the first two sentences, one number from the business's own books, one anonymized client moment, and one thing a competitor would not say out loud. A piece that clears all five goes on the belt. A piece that clears three goes back to the edit station.

How does the same content work for Google search and AI answers?

The same way, because both are looking for the same thing: a clear question, a direct answer in the first two sentences, and a passage that makes sense when lifted out of the page on its own. A piece built for a human who skims is already built for a machine that extracts.

The front door has moved, and the numbers say how far. Pew Research Center tracked the real browsing of 900 U.S. adults through March 2025 and found that about 18 percent of their Google searches produced an AI summary; when a summary appeared, users clicked a traditional result on 8 percent of visits, against 15 percent when it did not, and only 1 percent of visits included a click on a link inside the summary itself.††† Semrush's tracking of more than 10 million keywords put the share of queries triggering an AI Overview at 6.49 percent in January 2025, about 24.61 percent at the July peak, and 15.69 percent in November, with the mix shifting from 91.3 percent informational in January toward commercial and transactional queries as the year went on.§§ The share bounces. The direction does not.

What makes a page quotable by a machine is the same thing that makes it useful to a skimming human. Question headings, phrased the way a client asks. The answer in the first two sentences under each heading, before the story. Passages that make sense on their own, with no reliance on something said three screens earlier. A real author with a real name, a real firm and a real address, so the engine can resolve who is talking. Those are the house rules for every answer page worth building, and they are the reason a single page can rank for the human search, get quoted in the AI summary, and still hold its own as a thing a human would read start to finish. Getting named by ChatGPT and AI Overviews has its own post on SolvingHow.com; the point here is that the content system does not need a separate AI station. The edit station already builds the page the engines want.

One practical consequence for the ledger: write the question heading exactly as the client said it, not as a marketer would polish it. "Can my husband and I use the same divorce lawyer?" is the heading. "Joint representation in dissolution proceedings" is the heading a machine will never match to a human asking the first version at midnight.

How do small businesses measure whether content is generating leads?

By counting leads the content can be traced to, in a sheet the owner updates once a month, and by ignoring every other number until that one is answered. Likes, impressions and follower counts are chocolates in a hat: they make the station look busy and they have nothing to do with whether a piece got past the packing room.

The measurement is unglamorous and it takes 20 minutes. Every lead that comes in gets a source, captured two ways: a required "heard about us from" field on every form and intake script, and the page the lead first landed on, which Google Analytics or any form tool records for free. Once a month, in the last block of the month, the owner opens the sheet and writes four numbers next to each piece: visits, leads, clients signed, and revenue from those clients. That is the whole dashboard. A PE-backed platform with 11 branches runs the same sheet with a branch column. A regional 3PL runs it with a column for the sales rep who took the call, because the rep is the one who hears "I read the thing about pallet storage."

A worked example shows why this is the only sheet that matters. A two-partner family law firm publishes the Hillsborough County divorce cost page in January and one more answer page every month after it, so by December it has 12 pages and 12 ladders. The year-end sheet shows 2,400 visits to those pages, 41 intake forms that named one of the pages or came in directly from one, and 11 signed clients. That is a 1.7 percent visit-to-lead rate and a 27 percent close rate, both ordinary for a firm that answers the real question. The firm's average first-year value per family law client is $2,900, so the attributed first-year revenue is $31,900. The cost side is the owner's two-hour blocks, 52 of them, 104 hours, which at a $250 billable hour is $26,000 of opportunity cost, plus roughly $600 in tools: $26,600 all in. Year one clears by about $5,300 before counting referrals from those 11 clients. Year two is where it gets interesting, because the 12 pages are still live, still answering the question at midnight, and the only new cost is the next 12. That is not a hot take. It is arithmetic an owner can run on a napkin, and it is the arithmetic that decides whether the belt keeps running.

The Content Marketing Institute's 2026 survey found 33 percent of B2B marketers name measuring content effectiveness as a top challenge, and the report calls measurement the canary in the coal mine.‖ The canary dies first because most measurement starts with the easy numbers and never reaches the hard one. The sheet above starts with the hard one and refuses to look at anything else, which is why an owner can run it in 20 minutes and a marketing department often cannot run it at all.

What does a simple content marketing system look like step by step?

It looks like eight fixed stations, run in the same order inside a two-hour block, with one long piece a month and a ladder every week. Eight stations, one order, and a list an owner can tape above the desk.

• Step 1: Pick the home channel. The website, specifically answer pages, is the home for almost every service, professional and industrial business. Pick one delivery channel (LinkedIn for professional and industrial, Facebook or Google Business Profile posts for local service, Instagram for product businesses). Nothing else for six months.

• Step 2: Build the question ledger. One shared note. Seed it from 30 days of sent email, intake notes and call recaps. Log every new question within two minutes of the call ending, in the client's words, with the date. Target 40 or more questions on day one.

• Step 3: Block the two hours. Same afternoon every week, on the calendar as a client meeting, phone in a drawer. Partners alternate months. Coordinators own the stations; branch managers owe ten ledger questions a month.

• Step 4: Dictate, then draft. In block one of the month, pick the top question, dictate the answer into a phone the way it would be explained across a desk, and have an AI tool turn the transcript into a first draft with question headings and the answer in the first two sentences under each.

• Step 5: Run the human edit checklist. In block two, add one number from the business's own books, one anonymized client scene, one price range a competitor would hide, and one thing nobody else in the market says. Strip any sentence that could sit on another firm's site. The machine never writes prices, fees, legal claims, guarantees, statistics or citations. Read it aloud once.

• Step 6: Climb the ladder. In block three, cut the finished piece into one newsletter, five short posts (the number, the question, the scene, the mistake, the rule) and one 30-second video script. Every piece points back to the page.

• Step 7: Schedule a month of presence. Queue the five posts one a week, send the newsletter, record the video on a phone in one take. Weekly presence, monthly depth.

• Step 8: Run the sheet. In block four, update the measurement sheet: visits, leads, clients and revenue per piece, from the required source field and the landing page. Kill the formats that produce nothing. Double the ones that produce calls.

A two-partner law firm runs all eight with the partners alternating. A manufacturer runs them with the owner dictating and an office manager running stations four, six and seven. A solo consultant runs all eight alone, on Friday afternoons, and tells clients the autoresponder is not a bug. The stations do not change. Only the hands do.

Tocobaga builds this system for clients as a flat-fee engagement, never a percentage of anything, and the ledger, the sheet, the pages and every account they live in belong to the client from the first day, with no agency badge in the footer and no contract that renews itself. It is not the right fit for a business that wants volume for its own sake, and in that case the honest answer on the intro call is a referral elsewhere. For an owner who wants to see what the eight stations would look like for a specific firm, the fit call is at (813) 534-1261, hello@solvinghow.com or SolvingHow.com, and it is a conversation about the ledger, not a pitch.

Why do most small business content efforts fail after three months?

Because the belt was never designed, so it sped up until the owner quit. A content effort fails at month three for the same reason Lucy's shift fails in the first five minutes: there are no stations, every piece is an independent act of will, and will runs out.

The timeline is predictable enough to call in advance. Month one is enthusiasm and two pieces written in a burst of weekend energy. Month two is the first missed week, because a client emergency ate a block that was never really protected. Month three is the quiet decision not to catch up, and the blog's most recent date freezes there for a year, visible to every prospect who checks. The 13.9 percent of bloggers reporting strong results is not a measure of writing talent.‡ It is a measure of how many people built a line instead of a wrapping station.

The fixes are the ones already on the floor at Highland Park: a protected block, a full ledger, a machine on the draft station, a ladder, a sheet. Ford's workers were not stronger than Lucy. They had a line, and the line did not care how they felt on a given Tuesday.

The chocolate factory scene ends with the forewoman walking in on two women with full cheeks and lumpy uniforms, declaring them splendid, and calling for more speed. Nearly every owner who has tried to run content by hand has been in that uniform. The alternative was built in Highland Park a generation earlier, and it never required anyone to eat the chocolate: fix the order, bring the work to the station, and let the belt do the carrying. Ninety-three minutes a car. Two hours a week. Same idea.

What is a content marketing system for a small business?

A content marketing system is a fixed set of steps that turns customer questions into published pieces on a schedule, without the owner starting from a blank page each time. It includes one home channel, a running ledger of real client questions, a protected weekly time block, an AI-assisted first draft with a human edit checklist, a repurposing ladder that cuts each long piece into smaller formats, and a monthly sheet that tracks leads and revenue per piece.

How often should a small business publish blog content?

One substantial, specific piece a month is enough for most small businesses when each piece is cut into a newsletter, five short social posts and a short video that go out weekly. Orbit Media's 2026 survey found only 13.9 percent of bloggers report strong results, and its data ties strong results to effort per piece. Monthly depth with weekly presence beats daily thin posts.

Can a small business use AI to write blog posts?

Yes, for the first draft, with a human edit before anything is published. An MIT experiment on 453 professionals found ChatGPT cut writing time 40 percent and raised rated quality 18 percent, with the biggest gains for the weakest writers. The owner still adds the specific numbers, client scenes and price ranges, and never lets the machine write prices, legal claims, guarantees, statistics or citations.

How much time should an owner spend on content marketing each week?

Two hours, in one protected block on the same afternoon every week. Research from the University of California, Irvine found it takes an average of 23 minutes and 15 seconds to return to a task after an interruption, so a fragmented four hours produces less than a defended two. Across a month, the four blocks cover one long piece, its edit, its repurposing ladder and a 20-minute measurement update.

Where should a small business get content ideas?

From the questions clients ask on sales calls, intake forms and emails, written down verbatim within two minutes of the conversation. A single shared note seeded from the last 30 days of sent email usually yields 40 to 60 questions, which is years of monthly topics. Questions from paying customers are the only free keyword research that cannot produce a page nobody searches for.

How does a small business know if content marketing is working?

By tracking leads and revenue per piece, not likes or impressions. A required "heard about us from" field on every form and intake script, plus the landing page recorded by analytics, gives each lead a source. Once a month the owner logs visits, leads, signed clients and revenue next to each published piece, then cuts the formats that produce nothing and doubles the ones that produce calls.

Sources

*

Wikipedia. "List of I Love Lucy Episodes." Season 2, episode 1, "Job Switching," original air date September 15, 1952, episode 36 overall. https://en.wikipedia.org/wiki/List_of_I_Love_Lucy_episodes

†

Wikipedia. "Job Switching." Episode summary; ranked No. 2 on TV Guide's 1996 list of the 100 Most Memorable Moments in TV History and No. 1 among 30 moments in the 2013 Paley Center special TV's Funniest of the Funniest. https://en.wikipedia.org/wiki/Job_Switching

‡

Orbit Media Studios. "Blogging Statistics: The 2026 Annual Blogger Survey." September 8, 2026. Survey of 1,042 content marketers; average post takes 3 hours 20 minutes and runs 1,312 words; 13.9 percent report strong results; 92.4 percent use AI; frequency and effort trends since 2014. https://www.orbitmedia.com/blog/blogging-statistics/

§

Pattison, Kermit. "Worker, Interrupted: The Cost of Task Switching." Fast Company, July 28, 2008. Interview with Gloria Mark, University of California, Irvine; an average of 23 minutes and 15 seconds to return to a task after an interruption. https://www.fastcompany.com/944128/worker-interrupted-cost-task-switching

‖

Content Marketing Institute. "2026 B2B Content and Marketing Trends Report." October 2025. Survey of 1,015 B2B marketers fielded June 24 to August 14, 2025; 40 percent cite creating content that prompts action, 39 percent cite resource constraints, 33 percent cite measurement, 12 percent rate their content marketing highly effective. https://contentmarketinginstitute.com/b2b-research/b2b-content-marketing-trends-research

¶

Wikipedia. "Assembly Line." William "Pa" Klann and the Swift and Company disassembly line in Chicago; the Model T process divided into 45 steps; moving line began October 7, 1913, at Highland Park. https://en.wikipedia.org/wiki/Assembly_line

**

The Henry Ford. "Ford Methods and the Ford Shops." Flywheel magneto line from about April 1, 1913 (one unit per 20 minutes by hand, then one per 13 minutes, then one per 5); three continuous lines by early 1914 producing a Model T every 93 minutes versus 12.5 hours under stationary assembly. https://www.thehenryford.org/explore/blog/ford-methods-and-the-ford-shops

††

Wikipedia. "Highland Park Ford Plant." Moving assembly line from October 7, 1913; Model T price $700 in 1910 to $350 in 1917; $5 day announced January 5, 1914, raising the wage from $2.34 and cutting the shift from nine hours to eight. https://en.wikipedia.org/wiki/Highland_Park_Ford_Plant

‡‡

Soulo, Tim. "96.55% of Content Gets No Traffic From Google." Ahrefs, December 1, 2023. Study of about 14 billion pages in the Ahrefs index; lack of search demand named as the first cause. https://ahrefs.com/blog/search-traffic-study/

§§

Semrush. "AI Overviews Study: What 2025 SEO Data Tells Us About Google's Search Shift." Updated December 15, 2025. More than 10 million keywords tracked; AI Overview trigger share 6.49 percent in January, about 24.61 percent in July and 15.69 percent in November 2025; 91.3 percent informational in January; nearly 60 percent of triggering keywords have 100 or fewer monthly searches. https://semrush.com/blog/semrush-ai-overviews-study

‖‖

MIT Department of Economics. "Study Finds ChatGPT Boosts Worker Productivity for Some Writing Tasks." July 14, 2023. Summary of Shakked Noy and Whitney Zhang, "Experimental Evidence on the Productivity Effects of Generative Artificial Intelligence," Science, 2023; 453 college-educated professionals, time down 40 percent, rated quality up 18 percent, inequality between workers reduced. https://economics.mit.edu/news/study-finds-chatgpt-boosts-worker-productivity-some-writing-tasks

¶¶

Google Search Central. "Creating Helpful, Reliable, People-First Content." Self-assessment questions on original information, first-hand expertise, substantial value versus other results and search-engine-first content, including extensive automation across many topics. https://developers.google.com/search/docs/fundamentals/creating-helpful-content

***

Edelman. "Thought Leadership Gets B2B Buyers Back Into the Game." 2024 Edelman-LinkedIn B2B Thought Leadership Impact Report; 3,500 global B2B decision makers; 73 percent find thought leadership more trustworthy than marketing materials, 75 percent researched an unconsidered product because of it, 60 percent would pay a premium, 15 percent rate quality very good or excellent. https://www.edelman.com/insights/thought-leadership-gets-b2b-buyers-back-into-game

†††

Chapekis, Athena, and Anna Lieb. "Google Users Are Less Likely to Click on Links When an AI Summary Appears in the Results." Pew Research Center, July 22, 2025. Browsing data from 900 U.S. adults, March 2025; AI summaries on about 18 percent of searches; 8 percent versus 15 percent click rates; 1 percent click a link inside the summary. https://www.pewresearch.org/short-reads/2025/07/22/google-users-are-less-likely-to-click-on-links-when-an-ai-summary-appears-in-the-results/