Doing it yourself versus hiring: an honest cost comparison

The comparison is usually run as fee against zero, which is why so many owners start doing it themselves and stop by week six. Your time is not free and it is not evenly available. The hour you spend on profile photos in February is easy. The same hour in your busy season does not exist.

I ran restaurants. I know exactly what happens to the marketing task list during a bad week.

What does doing it yourself actually cost?

Four things, and only one of them shows on a bank statement.

Your hours, priced at what you would otherwise be doing with them. If those hours would have gone to hiring, sales, or fixing operations, that is the real cost.

Learning time, which is front loaded and larger than the doing time in the first months. This is a genuine investment and it does not disappear when you later hire someone, because it makes you a much better client.

Tools. Rank tracking, review management, and audit software all carry subscriptions. Some are worth it and several free options are adequate for a single location.

Error cost. Most mistakes here are cheap and reversible. A few are not, and they cluster around your profile identity and your domain. Those are the two places where a self-inflicted mistake can cost weeks.

The largest cost of doing it yourself is the version you start and abandon. A half-built review system that stopped asking in March is worse than never starting, because you believe it is running.

How do I put a number on my own hours?

Here is the arithmetic, and it uses federal data rather than a rule of thumb.

Start with the folklore, because you have probably been quoted it. Software and payroll blogs describe loaded labor cost as 1.25x to 1.4x base wage. No study sits behind that range. Every carrier of it cites another blog.

Now the real figure. The Bureau of Labor Statistics publishes Employer Costs for Employee Compensation. In the March 2026 release, accommodation and food services showed total compensation of $19.92 per hour worked against $16.12 in wages and salaries. That is a multiplier of 1.24. For an operator offering no health insurance, no retirement and no paid leave, wages plus legally required and supplemental pay comes to 1.13.

So where does 1.4 come from? The same table shows all private industry at $46.60 total compensation against $32.60 in wages, which is 1.43. Benefits are 30.1% of compensation economy-wide and 19.1% in food service. The folk rule imported an economy-wide number into a low-benefit industry. Use 1.13 to 1.24 for a restaurant, and pull your own figures if you are in a different sector.

Run it. If the person whose hours you would redirect is at $24 an hour, their real cost is roughly $27 to $30. Four hours a week for a quarter is 52 hours. At $28 loaded, that is about $1,456 of labor moved out of operations and into marketing.

Two honest limits. BLS wages exclude customer tips, and the industry code blends salaried managers with hourly staff, so it is not a line-cook number. And the multiplier prices a staff hour, not yours. Your own hour is priced by what it displaces, which is usually the highest-value thing on your list, not the cheapest. That is why owner-executed marketing is more expensive than it looks and better than it looks at the same time.

Which parts should you keep inhouse regardless?

The ones that need your knowledge of the business, and they are more valuable than owners assume.

Your primary Google Business Profile category. It is free, it takes ten minutes, and it is the single strongest lever in the list. It has ranked as the number one individual local ranking factor in every edition of Whitespark's expert survey, and a Search Engine Land analysis of 1.8 million Google Business Profiles across 4,209 categories found the exact-match category outranking adjacent ones for almost every category in the dataset. Nobody needs to be paid for this.

Photos. Real ones, of your actual space, staff, and work. Nobody outside your business can take these, and they matter.

Review asking. The request works best when it comes from the person who served the customer, in the moment. An agency can build the system, and the asking still happens at your counter.

Answering customer questions. You know the questions people actually ask because you hear them all day. Turning them into content is the highest-value writing available to a local business. The same knowledge feeds your profile, and seeding your questions section properly is a task an owner can do better than a vendor because you already know what people ask.

Hours, holidays, and anything about how the business runs. Nobody outside will catch a schedule change fast enough.

Which parts are worth buying?

The ones with a steep learning curve, a tooling requirement, or a real downside if done wrong.

Technical diagnosis. Why pages are not indexed, why a form is not tracking, why a migration lost positions. This is expert work and it is usually cheaper to buy a few hours of it than to spend a weekend guessing.

Anything irreversible. Site migrations, URL restructures, domain changes, reinstatement work on a suspended profile.

Production at volume. If your plan needs twelve pages, someone has to write twelve pages. That is a capacity problem and it does not get solved by learning.

Ongoing competitive monitoring. Not because it is hard, but because it is the first thing to lapse when you are busy.

Judgment about what is worth doing at all. Knowing which sources AI assistants actually pull from changes what belongs on your list, and that kind of filtering is often the most valuable hour you buy.

Which parts should you skip entirely?

The list of things not to spend a scarce hour on is worth more than the list of things to buy.

Conversion testing, at your traffic level. What a test requires is not a traffic figure, it is arithmetic you can do on the back of an envelope. Conversions needed per variant is roughly 16 ÷ r², where r is the relative lift you want to catch and the 16 is built from 2(1.96 + 0.84)², the 1.96 for 95% confidence and the 0.84 for 80% power. Run it backwards and the familiar hundred conversions per variant turns out to be exactly what detects a 40% lift (derived), and almost nothing you can change on a page moves conversion by two fifths. A local business at 300 to 800 monthly visits converting at 3% produces 9 to 24 leads a month, which splits into single digits per variant, so you would reach significance long after the season, the offer and the spend had all changed. Do the qualitative work instead, and measure before and after over long windows while saying plainly that it is not a controlled test.

Chasing AI search visibility as a local business, this year. Ahrefs analysed 146 million search results pages and found AI Overviews on 20.5% of them overall but only 7.9% of local searches. Sterling Sky independently observes AI-powered local packs on roughly 7% of tracked keywords. For one of Sterling Sky's largest multi-location clients, ChatGPT traffic went from 0.1% of Google traffic in 2025 to 2% in 2026, which is a big growth rate on a tiny base, and still only 22% of what Bing sends that client. Measuring it properly is also expensive in hours: the University of St. Gallen's work on AI answer stability recommends at least seven runs per prompt per day across a diverse prompt set, reported on a two to four week rolling window. That is a research program, not a task. Keep a claimed Bing Places listing, write for people, and revisit.

Geo-tagging your photos. In Whitespark's 2026 survey, 47 local search practitioners scored 187 factors, and presence of geo-tagged photos scored lowest of all of them. It is expert opinion rather than a controlled test. It is still the clearest available signal on a tactic that keeps being sold.

How do I actually decide?

Two numbers and one honest answer.

How many hours a week can you commit, every week, including your busiest month. Not your good weeks. Your average.

How many of those hours will go to learning versus doing, for the first quarter.

And the honest answer: have you started this before and stopped? If yes, twice, believe yourself. That is data.

Under about two reliable hours a week, do the human tasks yourself and buy the rest. At four or more, with genuine interest, a single-location business in a moderate category can run most of this alone and buy a few hours of review a quarter. That is not a reluctant admission. It is the correct answer for a large share of local businesses, and the category work above is the reason: the items that move the most are free and take minutes.

There is a middle option people forget. Pay someone for a defined diagnosis and a plan, then execute it yourself. You get the expertise where it is scarce and keep the labor where it is cheap. Ask for the plan as a document you own, which is a fair thing to buy on its own and looks like the discovery stage of a real engagement without the retainer attached.

One more variable, if you are considering hiring a marketing person instead. In accommodation and food services, Bureau of Labor Statistics turnover data for 2025 implies quits of roughly 50% of the workforce per year and total separations near 66%. Hiring for this role inside a high-turnover business means budgeting for doing it again, and the knowledge leaves with the person unless it is written down.

What to do this week

Track your marketing hours for one week. Actual hours, written down, not estimated. Most owners find a number lower than they expected.

Multiply those hours by your loaded rate using the 1.13 to 1.24 range, or your own sector's figure. Now you have a number to compare against a quote.

Then split your task list into three columns: only I can do this, anyone could do this, this needs an expert. The first column stays yours forever. The third is your shopping list. The middle column is where the decision actually lives, and it usually goes to whoever has the time.

If the middle column is large and your hours are small, price a scope for just that column. What a retainer should contain applies even when the scope is narrow, and a small honest scope is a legitimate purchase.

If you decide to hire, screen carefully, since finding a good search company is mostly a filtering exercise and you now know precisely which parts you want covered.

Questions about which parts to keep and which to buy?

Email me at eric@seod.com with the number of hours a week you can realistically commit during your busiest month, and your business type. I will send back which tasks to keep inhouse and which are worth paying for at that hour count, including when the honest answer is that you do not need to hire anyone. No pitch, no sequence after it.

There is more on hiring and working with agencies if you want to keep reading before deciding.