There's a specific hesitation we hear from first-time software buyers, usually phrased carefully: "does asking for an estimate cost anything?" It's a reasonable thing to worry about — nobody wants to accidentally trigger an invoice by asking a question. The honest answer is "it depends," and the more useful question underneath it is: what exactly are you trying to buy at this stage — a ballpark, or a plan you can actually fund?
Those are two different products, priced two different ways, and mixing them up is where a lot of buyer frustration comes from.
When the estimate should cost nothing
A ballpark range, delivered after a short call, should almost always be free. You describe your goals, your constraints, roughly what you're picturing. A serious partner responds with order-of-magnitude cost and timeline bands, plus a short list of what would move those numbers up or down.
That's genuinely enough information to decide whether a deeper conversation is worth your time this month. It is not — and this is the part people miss — enough to lock in a fixed contract for anything with real complexity. If a vendor offers to skip straight from that 20-minute call to a signed fixed-price agreement, ask yourself what assumptions they're quietly making on your behalf.
When paying for discovery is the smart move, not a red flag
Discovery becomes worth paying for once you need something more specific than a range:
- a prioritized MVP slice that actually reflects your real constraints,
- a technical risk review — integrations, data quality, AI evaluation, compliance,
- architecture options laid out with honest trade-offs, not just one "recommended" path,
- or a sprint plan detailed enough that you could hand it to a finance person and get a yes.
You're not paying for a PDF, even though a PDF might be the deliverable. You're paying to reduce the cost of being wrong before you're wrong at a much larger scale. A few thousand spent on discovery that reveals your "simple" integration is actually three integrations is a few thousand well spent.
An estimate that reads as one exact number with zero stated assumptions is a sales artifact, not a plan.
What a genuinely good estimate includes
We look for the same handful of things every time, regardless of who wrote it:
- Assumptions, written down — and specifically, what happens to the number if any of them turn out to be wrong.
- What's in scope and what's explicitly out for this first delivery. Not everything you'll ever want. Just this slice.
- The delivery model — sprints or fixed — and how often you'll actually see something working.
- Ownership terms — repository, cloud accounts, credentials, spelled out rather than implied.
- Quality defaults — tests, CI, staging, documentation — stated as included, not as line items you'll discover missing later.
- A range or a phased number, not false precision to the last dollar for something nobody has built yet.
If an estimate is missing more than one of these, that's worth asking about directly before you go further — not a dealbreaker necessarily, but a conversation.
Red flags worth taking seriously
- An instant fixed price offered after a fifteen-minute chat, for something that clearly touches multiple systems.
- No mention anywhere of who owns what when the engagement ends.
- "Unlimited revisions" dangled as a selling point, with no definition of done attached to it — that combination usually means scope creep is the business model.
- Pressure to sign before you've seen any example of how weekly delivery actually works with that team.
A sequence that tends to work well for buyers
- Free call — establish fit and get a ballpark. Fifteen to thirty minutes, no commitment either direction.
- Optional paid discovery — only if the range confirmed there's a real project here, and you need a specific first slice defined.
- First sprint — actual working software you're free to reject, extend, or take somewhere else entirely.
Notice that ownership doesn't have to wait until step three. It should be true from the start.
How to ask for an estimate without sounding difficult
Buyers sometimes worry that pushing for detail — assumptions, scope boundaries, ownership terms — will come across as high-maintenance before the relationship even starts. In practice, the opposite tends to be true. Vendors who do this well are usually relieved when a buyer asks sharp questions early, because it means less painful renegotiation later. The vendors who bristle at "can you write down your assumptions" are telling you something about how change requests will go mid-project.
A simple script that works: "Before we go further, can you send the assumptions behind this number, and what would change it?" That's not an aggressive question. It's the question a competent vendor expects and a weak one dreads.
What “free” sometimes really costs
There's a version of free estimates that isn't generous — it's a funnel. You get a polished number fast, with almost no discovery of your integrations or data. Then, once you've emotionally committed, the change orders begin. The free estimate wasn't free; it was the hook.
The tell is usually the missing assumptions page. A real free ballpark still says things like: "This assumes one core workflow, no legacy ERP sync, and a decision-maker available weekly." That sentence alone can save you a bad contract.
On the other side, some teams over-index and try to charge for every exploratory chat. For a simple fit conversation, that usually signals the wrong commercial posture for early-stage buyers. Fit calls should be cheap (ideally free). Decision-ready plans can be paid.
A note on AI-era estimates
In 2026, buyers hear that AI makes software "10x cheaper." Sometimes a slice really does move faster. Often the hard parts — messy data, permissions, integrations, evaluation — don't magically shrink. A good estimate separates:
- UI and workflow scaffolding that AI tooling can accelerate,
- from risk work that still needs senior judgment.
If a quote collapses those into one magical discount without naming the risks, treat it like any other unsupported promise.
ConaiSoft offers a short exploratory call at no charge, specifically to figure out fit and sketch a realistic first delivery. If your situation genuinely needs deeper discovery, we'll say so plainly — and we'll define exactly what that paid work produces before you spend anything on it.