Choosing a route
In house, agency, or an independent read first
Build in house when the capability is core to your business and you genuinely have the people free. Hire an agency when the requirement is settled and what you lack is capacity. Bring in a consultancy when the question is organizational rather than technical. Get an independent read first when you cannot yet say, with evidence, that the commitment is the right one.
The table below sets out what each route costs you, who owns the outcome, and what happens when the answer turns out to be wrong. The thing most comparisons skip is whether the decision is settled yet: if it is, this is a capacity question, and if it is not, you are about to buy a build to answer something a build cannot answer.
We do one of these four. The table says where the other three are better, because a comparison where one column wins every row is marketing, and you would be right to discard it.
| Your own team | Development agency | Management consultancy | Independent read first | |
|---|---|---|---|---|
| How fast you can start | Immediately, if the people are free. They rarely are, and pulling them off the roadmap has its own cost. | Weeks. Scoping, proposal and contracting before anyone writes code. | Weeks to months, and the engagement is usually larger than the question. | Days, because the scope is the decision rather than the build. |
| What it costs | Salaries you are already paying, plus the opportunity cost of what those people are not doing. Cheapest on paper, rarely cheapest in practice. | The largest number of the four, and the one most likely to grow. Integration and maintenance are commonly outside the quote. | High, and you are buying a recommendation rather than a working system. | The smallest of the four by a wide margin, and it is the only one that can end with you spending nothing else. |
| Who owns the outcome | You do, completely. There is nobody to escalate to when it slips. | Shared, and the boundary is whatever the contract says. Where several vendors are involved, often nobody. | Nobody. The deliverable is the advice, and execution is somebody else’s problem. | You do, which is the point. The output is a decision you can act on with or without us. |
| Depth of knowledge about your business | Total. Your team knows the history, the politics and which processes are fiction. | Shallow at the start and better by the end, though most of it leaves when the engagement does. | Built quickly and deliberately, then documented and handed over. | Enough for the decision in hand, and no more. That is a limitation, not a feature. |
| Incentive to tell you not to build it | Mixed. An internal team that wants the work is not neutral about whether the work exists. | None. They are paid to build, so the honest answer and the commercial answer point the same way only by coincidence. | Partial. They are not paid to build it, but they are paid for the next phase. | This is the whole job. The engagement ends either way, so a no costs us nothing and saves you the build. |
| What happens when the answer turns out to be wrong | You absorb it, and the people who made the call are the people who have to fix it. | A change request, usually chargeable. | A follow on engagement. | Nothing to absorb yet, because the commitment has not been made. That is the only reason to do this first. |
| Best when | The problem is well understood, the capability is core to your business, and you have the people. | The requirement is settled, the work is clearly scoped, and you need capacity you do not have. | The question is organizational rather than technical and spans several functions. | You are about to commit serious money and cannot yet say, with evidence, that it is the right commitment. |
Shaded cells mark where a route is genuinely strongest on that criterion. Three of the seven go to routes other than ours.
The row that decides it
Every route above except one is paid according to whether the work happens. Your own team wants the project, an agency is paid to build it, and a consultancy is paid for the phase after the recommendation. None of that makes any of them dishonest. It does mean that if you ask them whether to proceed, you are asking a question of someone whose answer costs them something.
That is the entire argument for doing a small, separate piece of work first, and it is the only row in the table where the fourth column wins outright. If your decision is already settled, it is not worth doing, and the other three columns are where you should be looking.
Questions people ask when choosing
- Should I build software in-house or hire an agency?
- Build in-house when the capability is core to your business, the requirement is well understood, and you genuinely have the people free. Hire an agency when the requirement is settled and what you lack is capacity. The honest warning on in-house is that the cost is hidden rather than absent: it is the roadmap those people stop working on. The honest warning on an agency is that a build quote typically covers the build and not the integration or the years of maintenance after it.
- What is the difference between a development agency and a consultancy?
- An agency builds what you ask for and is paid to build it. A consultancy analyses the problem and hands over a recommendation, usually without building anything. The practical consequence is that an agency has no commercial reason to tell you the project should not happen, and a consultancy has no obligation to make its recommendation work.
- Why would I pay for an assessment instead of just starting the build?
- Because the assessment is the cheapest of the four routes and the only one that can end with you spending nothing further. Published data shows projects that defined the requirement before committing land within roughly 10 to 15 percent of budget, while projects that skipped that step run 30 to 60 percent over. If the answer is that the build is right, you have lost very little. If it is not, you have avoided the whole number.
- Is an independent assessment just a sales process for the build?
- It is a fair question to ask of anyone who offers both. The test is whether the engagement is structured to end. Ask what the deliverable is if the answer is no, whether the fee changes depending on the recommendation, and whether you can take the output to a different firm. If the answers are vague, treat the assessment as a proposal.
Not sure which of these you are in?
Describe the decision you are facing and we will tell you which route fits, including when it is not us. Book a call or send a few lines, whichever you prefer.