Two proposals are sitting on your desk. One is a monthly agency retainer: a pod of three or four people, a sprint cadence, a deliverable list that lands every other Friday. The other is a fractional head of design: one senior person, a few days a month, no pod, owning direction and the quality bar. The retainer looks like more. More hands, more files, more visible motion for the money. My stance is that until your roadmap holds still for about a year, the single embedded leader returns more per dollar, because a seed to Series B company is almost never short on deliverables. It is short on decisions, and only one of those two proposals sells decisions.
Weigh that with the fact that I sell both. I run Surf Studios, where a multi-year engagement with Permatex has delivered 19 workstreams since 2024, and I also work with founders fractionally. I have also been the buyer on the other side of a retainer. At Enverus I hired and directed a web development agency to build the Deal Finder app I designed for NAPE, the largest show in oil and gas. So this is not an argument that agencies are a bad buy. It is an argument about which problem you have in front of you right now, and I will name the cases where I would tell you to sign the retainer instead.
What each proposal is actually selling
A retainer sells capacity. You are buying a set number of people for a set number of weeks, and the agency commits to cadence and volume. What a pod does not commit to is whether the work was worth doing. That part arrives from you, as a brief, a priority call, and an acceptance. A good pod will push back on a weak brief and improve it at the margins. It will not replace it, because nobody on that pod gets to decide what your company builds next, and they should not.
A fractional leader sells the opposite half. You are buying judgment, direction, and ownership of an outcome, priced in days per month rather than heads per sprint. Deliverables still show up. I still open the file and push pixels when the team is thin. But the deliverables are a byproduct of the calls, not the product you bought, and the person making the calls is inside your company, in your Slack, answering at the speed your questions actually arrive.
Now the line neither proposal prints: your own hours. A pod runs on briefs, and somebody senior has to write them, sequence them, review the output, and decide when a thing is done. At seed that person is the founder or the one product manager, which is the most expensive labor in the building. I watched the cost of a thin brief directly at the apparel platform where I lead design. An under-scoped handoff cost my team extra design cycles, so I built a more rigorous intake process that makes design ask its questions up front. That fix took a design leader to build. A pod cannot build it for you, because the questions it needs answered are the ones only you can answer.
A pod multiplies whatever brief you hand it. Before Series B, the brief is the part you do not have yet.
You are not short on output. You are short on aim.
Look at what happens to software output in aggregate. Pendo analyzed feature usage across 615 of its subscriptions and found that 80 percent of features in the average software product are rarely or never used (opens in new tab), and put the price tag on it: publicly traded cloud software companies invested up to $29.5 billion building them. Every one of those features shipped. Teams hit their dates. The throughput was never the failure.
The failure mode at your stage is sharper. CB Insights studied 431 venture-backed companies that shut down since 2023 and, of the 385 where a cause could be identified, 43 percent had poor product-market fit (opens in new tab). Running out of capital tops their chart at 70 percent, and the report is careful to call that the final cause of death rather than the root problem. Companies at your stage do not usually die with an empty backlog. They die having built the wrong thing on schedule.
This is why the cadence difference matters more than the headcount difference. Jeff Bezos put the standard plainly in Amazon’s 2016 shareholder letter (opens in new tab): "Most decisions should probably be made with somewhere around 70% of the information you wish you had," and "If you wait for 90%, in most cases, you’re probably being slow." Decisions at a seed company arrive hourly, not fortnightly. An agency pod meets that flow with a sprint ritual and a ticket, and every open question waits for the next call. A leader with context answers it in four minutes between other things, which over a quarter is the whole difference.
The roadmap test
Here is the test I would run on the two proposals before I read either price. Write down the design work for the next four quarters, today, and ask yourself whether you believe it. Not whether it is plausible. Whether you would bet the quarter on it. If you can name the surfaces, the sequence, and what done looks like, you have a stable roadmap and you should buy throughput. If the list dissolves the moment you imagine one customer conversation going sideways, you do not have a roadmap yet. You have a hypothesis with dates attached, and hiring a pod to execute it faster just gets you to the wrong answer sooner and with a bigger invoice.
The Permatex engagement is what a yes looks like. The brand is established, the catalog is known, the surfaces repeat: homepage redesigns, the mega menu, product listing pages, SEO, performance. Nineteen workstreams since 2024, each one with a shape you can see before it starts. That is a retainer earning its money, because the aim is settled and the only variable left is how much good work gets through the door each month. Nobody there needs me to decide what the company is. They need the work shipped at a bar, on a cadence.
The Deal Finder is what a no looks like, even though an agency built it. The decision was the product: a buyer enters their budget and what they are after, the system matches them to the vendors on that floor selling exactly that, and then it maps where to find them. It surfaced deals worth millions during the show. The agency I hired built that well, and they built it because I had already decided what it was. Hand that same pod a retainer and a vague ask for a show app, and you get something competent and wrong, invoiced monthly. If the deciding is the part that is missing in your company, buy the deciding first. That is the work I do as a fractional head of design: own the calls, set the bar, and be the person on the hook for the outcome rather than the file.
Where each dollar actually lands
The two invoices are not comparable line for line, so compare what the money touches. On a retainer you are paying a blended rate across a pod, which means part of every dollar buys senior attention and a larger part buys mid-level hands, account management, and the agency’s own overhead. That is not a scandal, it is how a firm stays solvent, and I price the same way. I broke the ratio down in what a product design firm costs. The number worth computing from any retainer proposal is how many senior hours per month are actually committed in writing, not implied by the logo on the deck.
On the fractional side you are paying for days per month of one senior person, with no benefits, no equity grant, no recruiter fee, and a ramp measured in days because they arrive senior. The rate per day looks high next to a blended pod rate, which is the comparison that misleads founders most often. I laid out the models in what fractional design leadership costs. Price the two against the same question: how much of this monthly number turns into decisions my company keeps?
Then price what survives the engagement. When a retainer ends you keep the files and, if you were disciplined about it, a system. The reasoning leaves with the pod. When a fractional engagement ends you keep the decisions in your team’s hands, a quality bar people can apply without you, and usually the hires who will carry it, because evaluating design candidates is part of the job. Both of those are real value. Only one of them keeps working after the last invoice clears.
When I would tell you to sign the retainer
Three conditions, and if you meet all three, the retainer is the better buy and I will say so to your face. First, the roadmap holds: you can write four quarters of work you believe in. Second, you already have someone senior inside who writes briefs and owns acceptance, whether that is a design lead, a strong head of product, or a founder with real design judgment and the hours to spend. Third, the bottleneck is genuinely hands. You have more defined work than your people can ship, and the fix is more people shipping it.
There is also a volume case that no single fractional leader can meet. The Enverus rebrand had to land across more than ten products and international brands at once, alongside a 90 by 40 foot double-decker booth for a show with more than 15,000 attendees. That is a quantity of production work that one person cannot physically produce, no matter how good the judgment. When you need a lot of defined output in a compressed window, you need a team, and renting one is faster than building one.
And the combination most founders miss is not either of these. It is sequencing. Buy the decisions first, then rent the hands against them. A fractional leader setting direction with an agency pod executing underneath is the arrangement I both run and work inside: at the apparel platform I direct external agencies and contractors alongside my own team of seven. A pod with a sharp internal owner is worth multiples of the same pod with a founder who reviews work between fundraising calls.
What to ask both proposals this week
Ask the agency who writes the brief, in writing, and what happens when it is thin. If the answer is a discovery phase, ask what you own at the end of it and who makes the call it recommends. Ask how a priority change mid-month is handled, and listen for whether the answer is a re-scope conversation or a change order. Ask what percentage of the retainer is senior time. A firm that is good at this will answer all four quickly, because they have had the conversation before.
Ask the fractional candidate what ships in the first 30 days, and push until the answer is a specific artifact with a date on it, not a plan to plan. I wrote down the version of that I hold myself to in the first 90 days as a fractional head of design. Ask who they would hire first and why. Ask what they will tell you no about. A leader you cannot imagine disagreeing with you is an expensive mirror.
Then run both answers through month seven, the month nobody designs for. A customer segment you thought was core turns out not to be, and half the roadmap moves. The retainer handles that with a scope conversation, a re-brief, and a few weeks of work already in flight that now has nowhere to go. The leader handles it by changing their mind on a Tuesday and redirecting the team Wednesday morning. If you expect month seven to happen, and at your stage you should, you have your answer.
The reason this call is worth an hour of your attention is that it is reversible but not free. A wrong retainer costs you two quarters and a backlog of files nobody asked for. A wrong fractional hire costs you less money and more time, because judgment you ignore is the most expensive thing on this list. Pick by the aim, not the headcount. If you want a straight read on which of those two proposals fits what you are building, tell me what is on your desk and I will tell you which one I would sign, including when it is the one I am not selling.