Most capacity decisions reach us in the same shape. The product has traction, the roadmap needs four engineers more than payroll has, and three proposals land on the CTO's desk with three different model names: staff augmentation, dedicated team, build-operate-transfer. The proposals quote different hourly rates, but the rate is the least interesting difference between them. The real question is what you own when the contract ends.
We run dedicated teams and BOT as core delivery models at SolutionPlus, and we take augmentation-shaped work when it is genuinely the right fit. So this is not a neutral overview. It is how the three models behave from the side that has to staff and govern them, including the cases where the model we charge the most for is the wrong one.
The three models in plain terms
Staff augmentation means individual engineers from a vendor join your team and work under your management. You assign the tasks and run the standups; process and architecture decisions stay yours. The vendor supplies people and replaces them when one leaves.
A dedicated team is a whole unit (engineers, QA, a technical lead) that works only on your product while the vendor runs delivery management. You set priorities and accept or reject the work; the team figures out how to build it.
Build-operate-transfer is a dedicated team with an end state. The partner builds the team, runs it on your product, then transfers the working capability into your organization: the engineers, the repositories, the runbooks, the operating habits. We published a week-by-week breakdown of the transfer phase separately. This article is about the choice between the three, not the mechanics of one.
The six dimensions buyers actually decide on
Rates dominate the first call, but the decision turns on six other things. Here is the honest grid, before the nuance. The ramp figures are internal planning ranges from our engagements; stack, hiring market, and access to your systems control the real schedule.
| Dimension | Staff augmentation | Dedicated team | BOT |
|---|---|---|---|
| Control | Full: your managers, your rituals | Shared: your priorities, vendor's delivery | Shared, shifting to you at transfer |
| Ramp time | Days to weeks per engineer | 4–8 weeks to a working team | 8–12 weeks, recruiting included |
| Cost curve | Flat, forever | Flat, lower per head than augmentation | Higher early, lower after a priced transfer |
| Knowledge retention | Leaves with the contractor | Sits inside the vendor's team | Moves into your org by design |
| IP | Yours by contract, scattered in practice | Yours, in one codebase and one team | Yours from day one, contractually and structurally |
| Exit cost | Lowest: end the contract | Medium: a handover period | High before transfer; low after, once transfer costs are settled |
Three of these rows deserve more than a cell.
Cost curve
Augmentation looks cheapest in month one and most expensive in year three, because the meter never stops and nothing amortizes. BOT inverts that: you pay build costs, recruiting, and governance up front, and the vendor's delivery fee typically ends at transfer if the contract prices it that way. What stays on your side of the ledger after transfer is payroll, any support retainer you keep, replacement hiring, and the transfer costs themselves. Our planning rule from the transfers we have supported since 2023: below twelve months of expected engagement, rent; past eighteen, owning starts to win. Between those marks, run your own numbers. Our rate band for dedicated and BOT work is $25–49 per hour depending on seniority and stack; a fully loaded senior engineer hired in-house in Germany runs €95,000–130,000 per year. We published the full 24-month comparison math in the BOT article.
Knowledge retention
This is augmentation's hidden tax. The contractor who held your payment integration together leaves on two weeks' notice, and the context leaves with them. A dedicated team concentrates the same risk rather than removing it: the knowledge stays current, but it stays with the vendor. BOT is the only model of the three where retention is a contractual deliverable rather than a hope, because the transfer phase exists to move habits, not just documents.
Exit cost
Augmentation's low exit cost is a feature, not a footnote. When the horizon is short or the roadmap is uncertain, cheap exit is exactly what you are buying. BOT's early exit is the most expensive of the three by design, because you are part-funding a capability you have not yet taken over. Anyone comparing models should price the exit, not just the month.
Where staff augmentation beats BOT
We talk a share of BOT inquiries out of the model, and most of them land here. Augmentation wins outright when the horizon is short: two engineers for four months to clear a backlog before a funding milestone will never amortize the transfer machinery (runbooks, ceremonies, hiring handoff, governance), so BOT's premium buys nothing. It also wins when you cannot commit yet. If a pivot is still ahead of you, locking a team shape into a transfer plan is premature; rent capacity, learn what the product is, then decide. And it wins when you already have strong engineering management. A VP of Engineering who wants hands on keyboards and has the rituals to absorb them gets little from a vendor's delivery layer, and paying for governance you will not use is waste. We say so on the call.
Where the dedicated team wins
Most of our clients land in the middle model, and the reason is usually management bandwidth rather than money. They want the product built and the priorities kept; they do not want a second job running standups and reviews for a remote team.
That is the shape of our dedicated delivery teams: German-led project leadership, a senior engineering team in Pakistan, delivery in English or German. Geography is the operating fact; governance is the product. The client keeps product ownership and architectural veto. We run the rest.
The dedicated model's honest weakness is dependency. The team is yours in every sense except the legal one, and if the vendor relationship ends, you keep the code but not the capability. Companies that feel that risk acutely are the ones BOT exists for. Companies that do not should not pay the BOT premium to insure against it.
Choose your model
- Choose staff augmentation if the horizon is under six months, your own engineering management has spare capacity, and the work is a spike rather than a product line.
- Choose a dedicated team if the roadmap runs for years, you want product control without delivery management, and you are comfortable with the capability living at the vendor.
- Choose BOT if you intend to own the engineering capability at the end, can commit to an 18–24 month arc, and want the cost curve to bend downward after transfer instead of running flat forever.
If you are scaling past the MVP stage and still split between dedicated team and BOT, our default advice is to start as a dedicated team with a priced transfer option written into the contract. The option costs little to draft and converts the BOT decision from a leap into a date you pick later. If you are split between augmentation and a dedicated team, the real question is who manages. Answer that honestly and the model picks itself.
If you want to pressure-test which model fits your roadmap, talk to us. The expensive mistake is signing a contract for one model while needing another, and that mistake is visible within the first month.
Frequently asked questions
Can we start with staff augmentation and switch to a dedicated team later?
Yes, and it happens regularly. It is cleanest when the same vendor runs both models, because the augmented engineers already know your codebase and can form the core of the dedicated team. Budget two to four weeks for the transition: delivery management has to be stood up, and priorities move from your task queue to a team backlog.
Is build-operate-transfer just expensive staff augmentation?
No. Augmentation has no transfer event: the engineers never become yours, and the vendor's commercial incentive runs toward keeping you renting. A real BOT contract contains a priced, dated transfer. Our planning rule is that below roughly twelve months of engagement, augmentation is the cheaper and simpler choice, because the transfer costs have no time to amortize, and we tell buyers so.
Who owns the code and IP in each model?
In all three, from the first commit, but only if the contract says so. Our contracts assign every line of code, design, document, and config to the client as it is created; not every vendor's do, so read yours before you sign. The real difference between the models is where the knowledge sits: with individual contractors, with a vendor's team, or inside your own organization after a transfer.
How do the rates actually compare between the models?
This is the first question on nearly every sales call. Our dedicated team and BOT work runs $25–49 per hour depending on seniority and stack. In the proposals our clients show us, German agency blended rates for senior work run €100–160 per hour, and a fully loaded senior in-house hire in Germany costs €95,000–130,000 per year. The model choice changes the shape of the spend more than the rate does.