Choosing a software development company Israeli businesses can rely on comes down to evidence rather than polish: a clearly written scope, honest communication, code ownership in writing, sensible security habits and a handover you can actually use. Judge every candidate on those five things, and a short scoring sheet, not on the size of the pitch deck.
Start with the problem, not the vendor
Most bad vendor relationships begin with a vague brief. Before you contact anyone, write one page that answers four questions: what business problem are you solving, who will use the software, what must work on day one, and what would make you call the project a success six months after launch.
This page does two jobs. It lets you compare proposals on equal terms, because every company is answering the same question. And it quickly shows you which companies ask good questions back. A partner who goes straight to a price without probing the problem is guessing.
Be honest about what you do not know. If you are unsure whether you need a custom system, an off-the-shelf product with integrations, or a small prototype first, say so. A good partner will help you narrow that down, and may even tell you that a smaller engagement is the right start.
Eight criteria for a software development company Israeli buyers can compare
These criteria separate partners better than portfolios or slogans do. Use the table as a reading guide while you review proposals and calls.
| Criterion | What good looks like | How to check it |
|---|---|---|
| Relevant experience | Has built something structurally similar, even in another industry | Ask them to walk through one comparable project and what went wrong in it |
| Communication | One named contact, a regular rhythm, written summaries | Watch how they handle your first week of questions |
| Scope discipline | Writes down assumptions and exclusions | Read the proposal for what it does not include |
| Engineering practice | Version control, code review, automated checks, staging environment | Ask them to describe how a change goes from a laptop to production |
| Security habits | Handles passwords, access and updates deliberately | Ask how they store secrets and who can reach your production data |
| Ownership and access | You own the code, accounts and domains | Check the contract and the account setup plan |
| Handover and documentation | Documents setup, deployment and decisions as they go | Ask what you receive at the end and in what form |
| Support after launch | Clear terms for fixes versus new work | Ask what happens when something breaks in month three |
Notice that language, location and hourly rate are not on the list. They matter, but they are tie-breakers. Time-zone overlap and the ability to work in Hebrew or English are practical questions to settle early; they do not tell you whether the work will be good.
Questions to ask in the first call
A first call is not a pitch to sit through; it is a chance to test thinking. These questions tend to surface real differences quickly:
- Who would actually work on my project, and will they be the people I speak to now?
- What is the riskiest part of what I have described, and how would you reduce that risk early?
- How do you handle a change request halfway through?
- Where will the software be hosted, and who will hold the hosting account?
- How do you test, and what do I see before anything goes live?
- Can I speak to someone for whom you built something similar? (Treat this as a reference check, not a formality.)
Listen for specifics. Answers like we follow best practices tell you little. Answers that describe a concrete practice, with the trade-off included, tell you the person has done the work.
Reading a proposal: scope, assumptions, exclusions
A proposal is only comparable if you read the same three sections in each one.
Scope
Is the work described as features and screens you can recognise, or as a vague paragraph? Good scopes name user types, main workflows and integrations. If a feature you consider essential is missing, ask whether it is included, not whether it was forgotten.
Assumptions
Every estimate rests on assumptions: that you supply content on time, that a third-party system has a usable API, that there is one round of design feedback. Assumptions are not a trick; they are how a vendor stays honest. The risk is when they are absent, because then nothing protects either side when reality differs.
Exclusions
Look for what is deliberately left out: data migration, content entry, training, hosting costs, licences for paid components, ongoing support. A low price with long exclusions is often more expensive than a higher price with a complete scope. For a deeper look at how pricing models shift risk between you and the vendor, read fixed-price versus time-and-materials contracts.
Code ownership, access and handover
This is where small decisions at the start prevent large problems at the end. Settle these points in writing before work begins, and have a qualified lawyer review the contract, since the details depend on your situation and on current rules.
- Who owns the code when the project is paid for, and what happens to third-party or open-source components inside it.
- Who holds the accounts: domain registration, hosting, source repository, app-store accounts, analytics and email services should be in your company's name, with the vendor added as a user.
- What you receive at handover: the repository with history, a setup guide, deployment steps, environment variable names, a list of external services and a short description of the architecture.
- Whether you could hire someone else to continue the work using only what has been delivered.
That last point is the real test. If the answer is no, you are locked in, whatever the contract says. The article on documentation and handover to avoid vendor lock-in goes through what a usable handover contains.
Red flags worth walking away from
- A fixed price and date offered after one short conversation, with no questions about your users or data.
- Reluctance to put ownership, hosting accounts or handover terms in writing.
- No staging environment, no version control, or the claim that testing happens after launch.
- The senior people appear in the sales meeting and then vanish.
- Every request is answered with yes, and no risks are ever raised.
- Pressure to decide quickly, or discounts that expire this week.
- References that cannot be reached, or work shown that they cannot explain in detail.
One red flag is a prompt to ask more questions. Several together are a reason to move on.
A simple scoring sheet
Once you have two or three proposals, score each against the eight criteria. Give each criterion a weight from one to three according to how much it matters to you, then score each company from zero to five. The numbers below are only an example of the method.
| Criterion | Weight | Company A score | Company B score |
|---|---|---|---|
| Relevant experience | 3 | 4 | 3 |
| Communication | 3 | 3 | 5 |
| Scope discipline | 2 | 5 | 3 |
| Ownership and handover | 3 | 4 | 2 |
Multiply each score by its weight and add them up. The total is a conversation starter, not a verdict. If the winner on paper feels wrong, find out why: the sheet may be missing something you care about, such as how well their team understands your industry.
Putting it together
Shortlist two or three companies, send them the same one-page brief, and compare their questions as carefully as their answers. If you are still deciding whether to use an outside team at all, start with outsourcing versus hiring in-house first, then return to this checklist. You can also see how one team describes its approach on the software development in Israel page, or look at the packages to see how engagements can be structured. Whichever route you choose, keep the written scope, ownership terms and handover list as your anchor.
Frequently asked questions
What should be in a software development contract?
At minimum: a described scope with assumptions and exclusions, how changes are handled, payment terms tied to deliverables, who owns the code and accounts, confidentiality, hosting and access arrangements, handover items and post-launch support terms. Have a qualified lawyer review it, because requirements and best practice depend on your circumstances and can change.
How do I verify a developer's past work?
Ask for a walkthrough of one comparable project, including what went wrong and how it was fixed. Request a reference you can contact directly, and try the live product if it is public. Ask who on their team did the work, since a portfolio from a former team member says little about your project.
Is a fixed price or hourly rate safer?
Neither is automatically safer. Fixed price suits a well-defined scope and shifts estimating risk to the vendor, but changes become negotiations. Hourly or time-and-materials suits evolving work and gives flexibility, but needs budget caps and regular reporting. Many teams use a fixed price for a defined first phase, then switch to time-and-materials.