What a useful Bristol software shortlist looks like
Search results for software suppliers tend to mix three different kinds of business together, and the mix is the reason so many buyers waste a first round of meetings. There are product engineering studios that build and maintain custom systems. There are managed service and infrastructure firms whose core trade is keeping networks, devices and licences running. And there are one or two person consultancies that subcontract delivery once the contract is signed. All three answer the phone to the same enquiry.
Before you approach anyone, decide which of the three you are actually buying. A shortlist that mixes categories produces proposals you cannot compare: one quotes a delivery team, another quotes a support contract, and the cheapest quotes neither. Our software development company directory lets you filter by what a supplier builds rather than by how loudly it markets, and the same discipline applies whether you found the name here or through a referral.
Bespoke build or configured product: settle this before the brief
A large share of enquiries that arrive at software studios describe a bespoke system when a configured platform would have done the job. The opposite mistake is just as expensive: forcing a process that is genuinely your competitive advantage into a template because a licence looked cheaper than a build.
The test is not complexity. It is whether the workflow is one your customers or regulators would notice if it changed. Invoicing, payroll, generic scheduling and standard commerce flows are solved problems with mature products behind them. Field data capture for a survey team, a pricing engine tied to your own risk model, a plant floor integration with instruments nobody else runs, a claims workflow shaped by a specific book of business: those resist configuration and will leak cost forever if you try.
A supplier worth shortlisting will say this out loud before you ask. If a software firm hears your outline and immediately scopes a ground up build without ever asking whether a product could cover most of it, treat the enthusiasm as a warning rather than a compliment.
Discovery is the phase that prices everything after it
Fixed quotes given before discovery are guesses dressed as commitments, and both sides pay for the guess later through change requests. A paid discovery engagement, run as its own small contract, is the single most effective way to make later proposals comparable.
Discovery should end with artefacts you own and can hand to a different supplier: a prioritised backlog written in your language, wireframes or clickable flows for the primary journeys, an integration inventory naming every system that must be read from or written to, a data model, a non functional statement covering expected load and availability, and a delivery plan with named risks. If what you receive is a slide deck and a total price, you bought a sales document.
Insist that discovery is separable. You should be able to pay for it, take the output and run a competitive process for the build without penalty. Suppliers confident in their delivery accept this readily. The ones who bury discovery inside an unbreakable programme are protecting a lock in.
Questions that actually separate suppliers in a proposal meeting
Generic evaluation criteria produce generic answers. These ones change the conversation:
- Who writes the code, and are they employees, contractors or a partner firm elsewhere? Ask for the split in writing, not a reassurance.
- Show me a repository handover you have done. Not a case study about launch, a story about the day a client took the codebase in house.
- What happens to the team if your largest client doubles its scope next quarter?
- Which parts of the estimate are you least confident about, and why?
- How do you handle the integration that turns out to be undocumented once you are inside it?
- What does your support arrangement cover after go live, and what explicitly falls outside it?
The fourth question is the most revealing. A team that cannot name its own weakest estimate either has not estimated seriously or is not being straight with you.
Commercial models and what each one is protecting
Fixed price protects your budget and transfers risk to the supplier, who prices that risk into the number and will defend scope aggressively. It suits work with a firm boundary, a replatform with known behaviour, an integration with a documented interface, a compliance driven change.
Time and materials protects the outcome and leaves budget risk with you. It suits genuinely new product work where the requirements will move, provided you get a real cadence of demos, a visible burn rate and the ability to stop at the end of any sprint.
A capped or target cost arrangement sits between the two and is underused. You agree a ceiling, you share the saving if the team comes in under it, and both sides have a reason to cut scope that is not earning its keep.
Retainers suit maintenance and steady evolution, not initial delivery. If a retainer is proposed for a build, ask what happens to the hours in a slow month and whether they roll forward.
Contract terms worth spending a meeting on
Intellectual property assignment should be explicit, should cover the work product and the deliverables from discovery, and should take effect on payment rather than on final acceptance of the whole programme. Ask directly about third party and open source components and how their licences interact with your product.
Source code access matters more than ownership on paper. Repositories, pipelines, infrastructure definitions and secrets should live in accounts you control from the first commit, with the supplier invited in. Anything else is an escrow conversation you will have under pressure.
Agree an exit in the contract while everyone is friendly: a notice period, a defined handover package, a rate for transition support and a commitment to documentation standards that make the handover possible. Add security expectations that match your sector, whether that is an ISO 27001 certified environment, penetration testing before launch or named restrictions on where data may be processed.
Local market realities in Bristol worth planning around
The engineering, aerospace, defence adjacent and financial services employers around Bristol compete for the same developers as the software studios you are interviewing, which pushes salaries up and keeps senior availability tight. Two consequences follow for a buyer.
First, a proposal promising a fully senior team starting immediately deserves a question about where those people are today. Second, blended delivery is common and normal here, with client facing leadership and architecture in the city and part of the implementation elsewhere. Blended is not a problem in itself, but it should be disclosed in the proposal rather than discovered in a standup, and the overlap hours should be written down.
Availability also moves with the academic calendar and with large programme cycles at the bigger employers. If your timeline is fixed, ask about start dates early, and be sceptical of a team that can begin next Monday with no explanation of what they just finished.
Turning the research into a comparison
Write the brief once, send the same version to every software supplier, and require the response in your structure rather than theirs. Ask each to price the same three phases so the numbers sit side by side, and ask each to name the assumption that would most change the estimate.
If a mobile client is part of the plan, compare specialists in mobile app development against the full stack studios rather than assuming one team covers both well. If the system will need to be found as well as built, keep technical search work in a separate conversation so it is not quietly dropped from a software build scope. When the brief is ready, request proposals from several suppliers at once and compare the answers, not the brochures.