Here is the number first, then the reasoning behind it. A working design studio spends somewhere between 3 and 12 percent of revenue on marketing, and most sit between 4 and 7. In monthly cash that is roughly $300 to $700 for a solo designer, $1,200 to $2,800 for a 2 to 5 person studio, $5,000 to $12,000 for a 6 to 15 person agency, and $15,000 to $40,000 once you pass 15 people. Those figures are cash only. Add founder time at an honest internal rate and almost every one of them doubles.
The percentage that actually applies to a design studio
The 3 to 12 percent band is a rule of thumb, not a measurement. It is worth knowing why professional services sit at the low end of it while product companies live at the high end.
Why services sit lower than products
A SaaS company can spend 30 or 40 percent of revenue acquiring customers because the marginal cost of serving the next one is close to zero and they stay for years. Spending a lot to win a subscription is arithmetic, not bravery.
You are not that. Every client you win consumes design hours you have a finite number of, and capacity is people, who are expensive and slow to add. If you doubled your spend tomorrow and it worked perfectly, you would be turning work away by Q3 or hiring in a panic, which is how studios ruin delivery quality. Studio marketing is capped by your delivery capacity, not by your appetite for growth, and that alone pushes the sane number into single digits.
The second reason: your work is your marketing
A logistics company has to buy attention because nobody sees a warehouse. You produce visible artefacts every month. A shipped site, a redesigned dashboard, a rebrand rolled out across a company. Each is a marketing asset the client has already paid you to make, and no other industry gets its raw material funded by its customers.
That is why studio budgets look small on paper and still work. The expensive part of the content sits in the delivery budget. What marketing pays for is turning delivered work into published proof, which costs far less than making it. A studio that publishes nothing is throwing away the cheapest input it will ever have, which is why the case study pipeline matters more than the ad budget at every size below fifteen people. Treat percent-of-revenue as a sanity check on all this, never as an instruction: it tells you when you are wildly out of line and nothing else.
The fixed floor every studio pays
Before any strategy, there is a floor. It exists at one person and at fifty, and it does not scale with headcount in any linear way.
What is on the floor
Domain, DNS and email, hosting and CDN for your own site, analytics, form handling or CRM, design tool seats used for marketing rather than client work, a scheduling link, a newsletter tool if you run one, subscriptions for whichever portfolio platforms you take seriously, plus the one or two directories that actually send you inquiries.
For a solo that is realistically $120 to $250 a month. For a mid-size agency it is $600 to $1,500, mostly because seats multiply and you start paying for a real CRM instead of a spreadsheet. It does not go to zero and it does not scale with revenue.
The line almost everyone forgets: your own site is a depreciating asset
Your studio site has a useful life. Ours is roughly two to three years before it starts misrepresenting what we do, and yours is the same whether you admit it or not. If a rebuild of your own site costs you $15,000 in internal time, and it lasts 30 months, that is $500 a month of real marketing cost you are pretending is free because no invoice arrives.
Treat it as a sinking fund. Put the monthly number in the budget from day one, so that when the site is three years old you have the money and, more importantly, you have already accepted that it is due. Studios that skip this end up apologising in sales calls for a portfolio site two positioning changes out of date, which is a much more expensive problem than the rebuild. The same logic applies to the real cost of a cheap website when the client is you.
What is not on the floor
Paid ads, sponsorships, awards entries, conference travel, and content production are all variable, and that is the point of separating them. When a dry quarter arrives you cut the variable layer and the floor keeps the lights on. If the fixed floor has crept up to 60 percent of your budget, you have no lever left to pull, and you find that out in exactly the month you need one.
Tier one: the solo designer
Assume $108,000 a year in revenue, which is $9,000 a month averaged out. A defensible band here is 3 to 7 percent of revenue in cash.
The worked example
Take 5 percent of $9,000. That is $450 a month.
Allocate it: $180 to the fixed floor (domain, hosting, email, analytics, one design tool seat attributed to marketing, scheduling), $100 to a portfolio platform and one community you are genuinely active in, and $170 variable, which buys one paid input a month such as a photographer for product shots, an editor for a case study, or a contest entry fee.
Now add time. Say 4 hours a week on writing, posting, editing case studies, and following up on old leads. That is about 17 hours a month. Priced at an internal cost of $85 an hour, that is $1,445 of time against $450 of cash. Total real spend: $1,895 a month, or 21 percent of revenue, of which 76 percent is time.
What that actually buys
Not much cash-wise, and that is correct. At this size your marketing is you: the work you publish, the people you already know, and the speed at which you respond. The $450 exists to remove friction, not to generate demand.
The highest-return use of a solo budget is production help, not distribution. Paying someone $150 to edit a case study you wrote badly at 11pm beats $150 of ads, because the case study keeps working for three years and the ad stops the day you stop paying. If you can afford one thing, make it the thing that improves the proof on your own site.
Where solos waste it
Three ways, reliably. Tools that automate a channel you have not yet made work manually. Directory listings that promise leads and deliver form spam. Small ad tests with no landing page behind them, which is buying traffic to a portfolio grid and then calling the silence a channel problem.
Tier two: the 2 to 5 person studio
Assume $420,000 a year, which is $35,000 a month. The band widens to 4 to 8 percent because you now have overheads that demand a more predictable pipeline.
The worked example
Take 5 percent of $35,000. That is $1,750 a month.
Allocate it: $400 fixed floor (more seats, a real CRM, better analytics), $350 to a freelance writer or editor working on two pieces a month, $500 to case study production including screen capture, photography, or short video, and $500 variable for experiments, sponsorships, or a paid channel test that you are willing to kill in eight weeks.
Founder time: 6 hours a week from someone whose billable value is $120 an hour is 26 hours a month, or $3,120. Loaded total is $4,870 a month, roughly 14 percent of revenue, with time still at 64 percent of the total.
What changes at this size
You stop being able to run marketing on remembering to do it. Two to five people means someone is idle when the pipeline dips, and idle designers are the most expensive thing a studio owns. The budget’s job shifts from removing friction to smoothing the curve.
The most valuable line in this tier is the $500 for case study production, because it converts work you already did into an asset that answers the question every prospect asks: can you do this specific thing for a company like mine. It is also where a properly structured landing page starts paying for itself, because you finally have enough traffic for structure to matter.
The honest risk
A lot of studios spend the variable half on visibility that flatters the founder and does nothing for the pipeline: awards, podcast appearances, community presence. Those pay out on a two year feedback loop, and a five person studio with three months of runway cannot make that its primary bet. Keep it to a slice, not the strategy. The test before you commit to any of them: name the kind of person at the kind of company you expect to reach, and say what you want that person to do next. If you cannot answer both, you are buying reputation. That is a fine thing to buy with a tenth of the budget and a bad thing to buy with half of it.
Tier three: the 6 to 15 person agency
Assume $1.5 million a year, which is $125,000 a month. Band: 5 to 9 percent, because at this size an empty month is not inconvenient, it is a payroll event.
The worked example
Take 6 percent of $125,000. That is $7,500 a month.
Allocate it: $1,200 fixed floor and platforms, $2,500 for a part-time content person or a contractor doing roughly 20 to 25 hours a month of writing and editing, $1,500 case study production including photography and video capture, $1,200 paid distribution and tests, and $1,100 amortised across events, awards, and sponsorship commitments that get paid in lumps.
Time: a partner at 8 hours a week (35 hours at an internal value of $150 is $5,190) plus a designer at 4 hours a week (17 hours at $70 is $1,211). That is $6,400 of time. Loaded total is $13,900 a month, about 11 percent of revenue.
Notice what happened between tier two and tier three. Cash went up 4.3 times, time went up about twice, and the time share fell from 64 percent to 46 percent. That crossover is the real definition of a studio that has built a marketing function rather than a marketing habit.
What the number buys that smaller budgets cannot
Consistency and compounding. At $7,500 a month you can publish on a schedule instead of in bursts and maintain a body of work that ranks and gets cited. You can also afford to lose money on a channel for two quarters while you learn whether it works, which is the difference between a studio that has a channel and one that has anecdotes.
This is where the technical side of your own site starts to matter commercially rather than professionally. Slow pages and broken structure cost you inquiries at this volume, and Core Web Vitals on your own domain become a line item rather than a hobby.
The pitch trap at this size
Free concept work. At this size you get invited into competitive pitches, and speculative design is tempting. Do the arithmetic first: three designers for a week is easily $8,000 to $12,000 of internal cost, which is more than your entire monthly marketing budget spent on one prospect with a one-in-four chance. If you do it anyway, book it as marketing spend and cap the number of pitches per quarter. When we started putting ours in the marketing column, the number we entered dropped inside two quarters, because a line item gets argued about and a Friday night does not.
Tier four: fifteen people and up
Assume $4.5 million a year, which is $375,000 a month. Band: 5 to 10 percent, and this is where the top of the band is a deliberate choice rather than a symptom of panic.
The worked example
Take 6 percent of $375,000. That is $22,500 a month.
Allocate it: $8,000 fully loaded for one in-house marketing lead, $4,500 for a contractor pool covering writing, editing, and video, $2,000 tooling and platforms, $3,000 paid distribution, $3,500 for events and sponsorships amortised monthly, and $1,500 into the site and brand sinking fund.
The interesting line is the first. $22,500 a month is roughly the threshold where a full-time marketing person stops being a luxury: they cost about a third of the budget and there is enough left for them to execute. Below it you are hiring someone and then not funding their work, which is the most common way studios conclude that marketing hires do not work.
Category spend versus pipeline spend
Past fifteen people, part of the budget stops being about next quarter. Sponsoring a conference, producing a research piece, or building a tool your peers use are category plays: they change what people think you are, over years. They are legitimate, and they are also where budgets get burned with no accountability.
Split the number explicitly. We would put 70 to 80 percent on pipeline work you can trace to inquiries within two quarters, and 20 to 30 percent on category work judged annually on different criteria. One pot means the slow work gets cancelled every time a quarter goes badly, which guarantees you never find out whether it worked.
Growth mode and maintenance mode are different budgets
The percentage bands above are steady state. Real studios spend most of the year in one mode and a couple of quarters in the other, and the numbers are genuinely different.
Maintenance mode
Pipeline is healthy, delivery is at 80 to 90 percent capacity, and the goal is to not go backwards. Spend at the bottom of your band: 3 to 5 percent. The job is publishing what you ship, keeping the site current, and staying in touch with past clients. You are protecting an asset, not building one.
Maintenance is not a failure state. A studio at capacity spending 4 percent for eighteen months is doing something right. The mistake is confusing maintenance with neglect: it still means publishing, because the lag between publishing and inquiry is long enough that stopping is a decision about next year, not this month.
Growth mode
You are opening a new service line, changing positioning, entering a new market, or you have just added capacity you now have to feed. Spend at the top of the band or above it: 8 to 12 percent, deliberately, for two to four quarters, with a written end date.
The end date matters more than the amount. Growth spending without a review date turns into your new baseline, and then a soft quarter arrives and you cut it in a hurry, which wastes everything the ramp had built. Decide the duration up front, decide what evidence would make you extend it, and put both in writing where your business partner can see them.
The mode you must not stay in
Reactive mode: nothing when busy, then heavy spending when the pipeline empties. Studio marketing lags by roughly three to nine months between effort and signed work, so panic spending always arrives a quarter too late to fix the quarter you are panicking about.
Your first hire is almost never a marketer
Below roughly ten people, hiring a marketer is usually the wrong call. The salary is real, the budget around it is not, and the thing you are short of is rarely promotion.
What that salary is competing with
A marketer at $4,000 a month costs $48,000 a year. In a six person studio that is the same money as a mid-level designer who directly increases what you can deliver and, indirectly, what you can publish. The marketer’s output depends on inputs you cannot yet supply: case studies nobody has written, work nobody has photographed, a positioning statement the founders have not settled.
Marketers do not create the raw material, they amplify it. Hiring one before the raw material exists produces a lot of posting, some brand-voice documents, and no pipeline, after which everyone concludes marketing does not work for design studios.
The hire that usually beats it
A producer. Someone who writes case studies, captures screens and photography properly, runs the publishing schedule, and chases the details founders keep dropping. This role converts existing work into assets, which is the actual bottleneck, and it costs less than a marketing lead. Second best is a part-time senior doing two days a month on strategy with a producer executing: you get the judgment without funding a full seat.
The real exception
If the founder is the only person who can sell and the founder is at capacity, a hire that buys back founder hours is correct even at a small size. Sometimes that is a marketer; more often it is a project lead who takes delivery off the founder’s plate. The test is not the job title, it is whether the hire frees the person whose face and judgment generate the work. Studios weighing this are usually also weighing what separates a freelancer from an agency in their clients’ eyes, and the answer is mostly whether the founder is a bottleneck.
What counts as marketing spend, and what people wrongly leave out
Every studio we have compared notes with underreports its marketing spend, usually by around half, and then wonders why the return calculations feel wrong. Here is what belongs in the number.
Founder time, priced honestly
Not at your client rate, which overstates it, and not at zero, which is what most people do. Use internal cost per hour: total compensation plus overhead divided by realistic working hours, or 40 percent of your billable rate as a stand-in. Then log marketing hours for one month, honestly, including the ones spent on conversations that turned into a referral. The number will be larger than you expect and it will change what you decide.
Case study production
A proper case study takes 6 to 10 hours of writing and structuring, 4 to 8 hours of visual production, and a round of client approval. At an internal cost of $60 an hour that is $600 to $1,080 each. Six a year is $3,600 to $6,500, which is $300 to $540 a month that most studios book as zero because it happens on a Friday afternoon.
Budget it as a line and it gets done. Leave it implicit and it happens twice a year, badly, and your portfolio ends up representing what you were doing eighteen months ago.
Portfolio photography and product capture
Screens do not photograph themselves. Good product capture means undistorted device mockups, screenshots taken at the right pixel density, video of real interactions rather than a Figma prototype, and photography for anything physical. Budget $300 to $1,500 per project depending on whether video is involved. This is the most commonly skipped line, and the one that separates studios whose work looks credible online from studios whose work is just as good and does not look it.
The discounts you give to win the logo
If you sell a $20,000 project for $12,000 because you want that client in your portfolio, you just spent $8,000 on marketing. It is a legitimate spend and it is invisible unless you write it down. The same goes for pro bono work, speculative concepts in pitches, and referral commissions, which are marketing spend rather than a cost of delivery no matter how your bookkeeper files them.
Track these for a year and the arithmetic moves on its own: add the discounts, the spec work and the founder hours to a 3 percent cash budget and the honest total routinely lands nearer 9, with most of the difference going somewhere nobody chose.
What is not marketing
Proposal writing after a qualified conversation is sales. Account management is delivery. Keep them out, not because they are cheap but because mixing sales cost into marketing cost makes both numbers useless. What an inquiry costs to turn into a signed client is a separate calculation with separate inputs, and it stops being answerable the moment proposal hours are sitting inside your marketing total.
Moving the number with your pipeline
The budget is not a fixed monthly bill. It should move, but on evidence rather than on mood.
The signal to use
Pipeline coverage: the weighted value of qualified opportunities for the next quarter divided by the revenue you need in that quarter. If you need $105,000 booked and you have $60,000 weighted, your coverage is 0.57, which is a growth-mode signal regardless of how busy this week feels.
Below 1.0, move to the top of your band. Between 1.0 and 2.0, hold. Above 2.0, either invest the surplus in capacity or accept that you are about to have a scheduling problem. This one ratio beats any percentage, because it responds to reality with a quarter of lead time instead of reporting last month’s mood.
Spend before you need to, not after
Because of the three to nine month lag, the right time to increase spend is when you are comfortably busy and your coverage is starting to thin. That feels wrong: you are at capacity, the money is coming in, and paying to win work you cannot yet deliver sounds like a mistake.
It is not. You are buying pipeline for two quarters out, when the current projects have shipped. The studios that never have a dry quarter are the ones that spend into a full pipeline, and the ones that lurch from feast to famine are spending in arrears.
What to cut, and in what order
When you genuinely must cut, cut in this order: paid distribution first because it stops instantly and restarts instantly, then sponsorships and events because they are lumpy, then contractor content production, then the fixed floor, which you should almost never touch. Do not cut publishing entirely. Reduce the cadence, keep the habit, and keep measuring what the site actually does so you know what you restarted when things recover.
The trap of percentage-of-revenue when revenue is lumpy
This is the part that breaks the neat arithmetic above, and it deserves its own treatment because it is the failure mode we see most often.
Why the monthly percentage is nonsense
A four person studio bills $12,000 in January, $8,000 in February, $61,000 in March because two milestones landed, and $9,000 in April. Five percent of each month is $600, $400, $3,050, $450. No marketing plan survives that. You cannot hire a contractor for one month, you cannot run an eight week test on a budget that swings sevenfold, and you will always be cutting in the months your pipeline needs you most.
Use trailing twelve months, and skim at the invoice
Sum the last twelve months of revenue and divide by twelve. If that is $30,000, your 5 percent is $1,500 a month, every month, regardless of what arrives this week. Recalculate quarterly, not monthly.
Then make it real. Skim the percentage off each payment as it lands and move it into a separate account. When a $61,000 month arrives, $3,050 moves across, and it is still there in February when nothing came in. Ten minutes of setup, and the February budget stops depending on whether February had an invoice in it. It is the same discipline that makes retainers easier to run than one-off projects, applied to your own business instead of a client’s.
Handle the concentration risk separately
If one client is 40 percent of your trailing revenue, your percentage is built on sand. Set the budget on revenue excluding your largest client and treat anything above that as surplus for growth mode. It sounds conservative. What it means is that when the big client pauses, and eventually one always does, your marketing does not stop in the same month your revenue does.
How to choose your number
Benchmarks are for arguing with your business partner. Here is a procedure you can run this week and finish in about two hours.
Step one: establish the two baselines
Calculate trailing twelve month revenue divided by twelve. Then audit your last 90 days of actual marketing spend, including founder time at internal cost, case study production, photography, discounts given for portfolio value, and any spec work. Divide by three. Compare the two as a percentage. Most people are surprised in one direction or the other, and either surprise is useful.
Step two: decide your mode, on the ratio not the vibe
Compute pipeline coverage for next quarter. Below 1.0 is growth mode. Between 1.0 and 2.0 is maintenance. Above 2.0, your constraint is capacity and this article is not your problem. Write the mode down with the date, because you will want the history in six months.
Step three: pick the percentage and subtract the floor
Maintenance takes the bottom of your tier band, growth the top or slightly above. Apply it to the trailing average, then subtract your fixed floor including the sinking fund for your own site. What remains is your only actual decision, and it is much smaller than the headline number, which is exactly why the headline number causes so many bad choices.
Step four: allocate the remainder to production first
Fund the assets before the distribution: case studies, photography, the writing, the parts of your site that carry the argument. Distribution amplifies whatever exists, and if what exists is thin you are paying to show more people something unconvincing. If money is left once production is funded, test one channel the way you would run an A/B test on a page: one variable, a fixed window, a number you agreed to hit before you started, and a decision at the end of it.
Step five: set the review date and the change rule
Quarterly review. Change the number only on a two-quarter trend, never on one bad month. Write down in advance what evidence would make you increase it and what would make you cut, because deciding that mid-crisis produces the reactive pattern this article is trying to talk you out of.
Step six: decide what you will stop
New spending without stopping something is how a 5 percent budget quietly becomes 11 percent of a smaller revenue base. Every quarter, name one thing you are ending: the directory that sends nothing, the award you enter out of habit, the channel you have been giving another three months to for the last eighteen.
If you run all six steps and end up close to where you started, the exercise still paid for itself, because you now know the number instead of inheriting it. Then spend the first month of the new budget on your own site rather than on distribution: the about page and the words on the work usually move more inquiries than an extra thousand a month in ads, and they keep working after you stop paying.
If you are working through this for your own studio and want to compare notes, or want a second opinion on where your budget is going, write to us at hello@beconfidency.agency or through the contact page. Bring your trailing twelve month average and your last 90 days of actual spend, founder hours included, and we will run the six steps against real numbers instead of a benchmark.
If you would rather hand this whole discipline to one accountable team, that is exactly what our web design service is for.