Between roughly 5 and 25 people, most startups hit the same gap. The founder has been the de facto product manager since day one, and it worked, because the product fit in one head. Now there are eight engineers, two big clients with opinions, and a backlog that lives in Slack threads. The founder is doing product leadership in the gaps between fundraising and sales, which is to say: nobody is doing it.
The textbook answer is "hire a head of product". The textbook is expensive. A good senior product hire in South Africa runs R1.2m–R1.8m a year, takes three to six months to land, and is brutal to get wrong at this stage: a mis-hire doesn't just cost the salary, it costs a year of product direction.
Fractional product management exists for exactly this gap.
What fractional actually looks like
Not advice. Not a weekly call where a consultant reviews slides. A working fractional PM embeds two to three days a week and owns things:
The backlog and specs: engineers get a single accountable answer to "what's next and why", with acceptance criteria instead of Slack archaeology. The discovery cadence: customer interviews keep happening even when sales is busy, so the roadmap stays connected to reality. The delivery rhythm: planning, refinement and retros that produce decisions, run consistently enough that the team stops relitigating process. And the stakeholder line: someone whose actual job is to say "not now" to the big client, with a reason attached, so the founder doesn't have to spend goodwill doing it.
The two-to-three-day constraint is a feature, not a compromise. It forces systems over heroics: everything gets written down, every process has to survive the days the fractional PM isn't there. Which is exactly the machinery you'll hand to a full-time hire later.
Honest numbers
A fractional retainer at two days a week typically runs 35–50% of a full-time senior product salary, with no equity, no notice-period risk, and a start date measured in days. For international readers: South African senior product talent operates at a meaningful discount to UK/EU rates with full timezone overlap with Europe, which is why a growing share of fractional work here serves offshore scale-ups.
The right comparison isn't "fractional vs. full-time salary", though. It's "fractional now vs. six more months of nobody doing the job". Price the second one in shipped-the-wrong-thing terms and the retainer looks cheap.
When fractional is wrong
Honesty requires the other column. Fractional is the wrong call when product is the company's only job and the founder should be doing it themselves; when you need someone in the building five days a week managing people, not product; or when the codebase is the bottleneck, not the direction. That's a CTO problem wearing a product costume.
And a good fractional engagement should end. The healthy pattern is 6–12 months: install the systems, raise the bar, help hire the full-timer, hand over, leave. If your fractional PM isn't actively trying to make themselves unnecessary, you've hired a dependency, not a leader.
The test
If your engineers can't tell you what they're building next week and why it matters, you're in the gap. The only question is whether you close it with a six-month hiring process or a two-week start. That's the trade fractional makes, and for most teams between 5 and 25 people, it's not close.