Hiring guide
Software consultant or software developer? What each one is for
By Samuel Marndi · Updated
The short answer
A developer builds what you specify. A consultant works out what to specify. If you know exactly what you want built, hire a developer and do not pay consulting rates for it. If you are about to spend serious money and you are not certain the plan is right, a few days of consulting is the cheapest insurance you will ever buy.
The expensive version of this mistake is hiring a builder for a decision. Ask a developer whether you should build something and the answer is usually yes, because building is what they do and it is not dishonest of them — it is what you asked. A consultant is paid to be able to say "buy this instead, it costs a tenth as much", and to still be paid for the day.
The five jobs a consultant does that a developer does not
| Job | What it produces | Typical length |
|---|---|---|
| Feasibility and scoping | A written scope you can hand to any developer and get comparable quotes against | 2 – 5 days |
| Build versus buy | A costed comparison of building, buying, configuring an existing product, or doing nothing | 2 – 4 days |
| Architecture and stack choice | A decision with its reasons written down, so the next person understands why | 3 – 8 days |
| Audit of what you already have | A findings document: what is at risk, what is wasted, what to fix first, with costs | 3 – 10 days |
| Rescue of a stalled project | An honest assessment of what is salvageable, and a plan that does not require starting over | 1 – 3 weeks |
Which one your problem needs
| If this is your situation | You need | Because |
|---|---|---|
| "I need a website with these eight pages" | A developer | The outcome is describable; paying consulting rates to describe it again is waste |
| "Our processes are on WhatsApp and spreadsheets and it is breaking" | A consultant first | The answer may be software, or it may be configuring something that already exists |
| "We were quoted ₹18 lakh and we cannot tell if that is right" | A consultant, for two days | A second opinion on a large quote is the highest-return money in this list |
| "Our app is slow and we do not know why" | A consultant, then a developer | Buy the diagnosis, then buy the fix at a fixed price against it |
| "Our previous developer left and nothing is documented" | A consultant | Someone has to work out what exists before anyone can safely change it |
| "Our cloud bill tripled" | A consultant | Almost always configuration and waste, not capacity — and a bill is diagnosable |
| "We know what to build and have the designs" | A developer | Hire the builder, and hold them to the written scope |
What an audit actually finds
This is the most common engagement and the most consistently surprising one, because the findings are rarely what the client expected. The pattern across small and mid-sized Indian businesses is remarkably consistent.
- Accounts in the wrong name. The domain registered to a former developer's email, hosting on an agency's account, the Play Console under someone else's login. Usually fixable, occasionally not, and always worth finding before it matters.
- Backups that are not backups. A copy on the same server, or a RAID array mistaken for a backup — see why RAID is not a backup. Nobody has ever tested a restore.
- Cloud spend on things nobody uses. Idle instances, orphaned disks, over-provisioned databases, a staging environment running production-sized machines since 2023.
- One person who knows everything, nothing written down, and no second pair of eyes on anything that matters.
- Software bought and never rolled out. Licences renewing annually for a tool three people tried once.
- Security basics missing — default credentials, no multi-factor on administrative accounts, leavers who still have access. The small-business security checklist covers what to fix first.
- And frequently: the thing being built is not the thing needed. Which is the finding that saves the most money and is the least comfortable to deliver.
Build versus buy, done properly
This decision is usually made on instinct and then defended for years. Done properly it takes a couple of days and compares four options honestly — including the two nobody puts on the list.
The two missing options are "configure something that already exists" and "do nothing and fix the process instead". A surprising share of custom software requests are a process problem with a technology budget attached, and software built on top of a broken process reliably makes it faster and worse.
| Option | Costs most in | Right when |
|---|---|---|
| Build custom | Money upfront and maintenance forever | The process is genuinely your competitive advantage, or nothing on the market fits |
| Buy off the shelf | Monthly licence fees and fitting your process to the tool | Your need is ordinary — accounting, CRM, payroll, helpdesk. Most needs are ordinary |
| Configure and integrate | Integration work, and a dependency on vendors' APIs | Two or three existing tools nearly do it and the gap is the joining up |
| Do nothing, fix the process | Management effort and uncomfortable conversations | The problem is agreement, not tooling. More often than anyone wants to hear |
How a consulting engagement should run
Short, fixed, and ending in a document you own. If a consulting engagement has no defined end, that is not consulting — it is an unstructured retainer, and you should buy it knowingly or not at all.
- A scoping call first, free. If the answer is "you do not need a consultant for this, you need a developer", that should be said on the call and cost you nothing.
- A fixed number of days, agreed upfront, with what will be examined and who will be spoken to.
- A written findings document, addressed to a decision-maker rather than an engineer, with costs against each recommendation and a stated order of priority.
- A walkthrough, because nobody acts on a document they have only been emailed.
- No obligation to use the same person to implement it. A consultant whose recommendations only ever conclude "hire me to build this" is a salesperson with a nicer invoice. The findings should be hand-able to any developer or agency.
What it costs, and why it is the cheapest line in the project
Consulting is bought in days, and in India an independent consultant's day rate sits well below a firm's. The reason it pays for itself is arithmetic rather than optimism: three days of scoping against an ₹18 lakh build is under 2% of the budget, and the most common finding is that the build should be materially smaller or shaped differently. On a quote that size, being 20% wrong about the scope costs more than the entire engagement.
My own ranges are on the pricing page, and the deeper strategy, procurement and roadmap work sits under IT consulting. For the discipline-specific versions: cloud consulting for bills and migrations, AI consulting for what is worth building with AI and what is a demo, and custom software once the decision to build has survived scrutiny.
The honest routing, including the cases where you do not need either.
FAQ
Related questions
What does a software consultant actually do?
Works out what should be built, and whether it should be built at all. In practice that is five jobs: scoping a project so you can get comparable quotes, deciding build versus buy, choosing an architecture and recording why, auditing what you already have, and rescuing stalled projects. The output is a document you own and can hand to any developer — not code.
What is the difference between a software consultant and a software developer?
A developer builds what you specify; a consultant works out what to specify. Ask a developer whether you should build something and the answer is usually yes, because that is what you asked them about and building is what they do. A consultant is paid to be able to say "buy this instead, it costs a tenth as much" and still be paid for the day.
Do I need a consultant if I already know what I want?
No, and you should not pay consulting rates to have a known outcome described again. Hire a developer and hold them to a written scope. The exception is size: on a quote above ₹10–15 lakh, two days of independent review is under 2% of the budget and the most common finding is that the scope should be smaller or shaped differently.
How much does a software consultant cost in India?
Independent consultants are bought in days, and an independent day rate sits well below a consulting firm's. The useful way to judge it is as a share of what it is protecting: three days against an ₹18 lakh build is under 2% of the budget. My own ranges are on the pricing page, with a written itemised quote after a free scoping call.
What is a software audit and what does it find?
A structured review of what you have, producing a findings document with costs and priorities. The recurring findings across Indian SMBs are remarkably consistent: accounts registered in a former developer's name, backups nobody has ever test-restored, cloud spend on idle resources, one person who knows everything with nothing documented, software licences renewing for tools nobody uses, and missing security basics. Frequently also that the thing being built is not the thing needed.
Can I hire a software engineer for just a few days?
Yes — short fixed engagements are the normal shape for diagnosis, review and scoping work, and the better arrangement when nobody can scope the full job yet. Buy a few days, get a written finding, then buy the implementation at a fixed price against it. What you should avoid is a consulting engagement with no defined end, which is an unstructured retainer under a different name.
Should the consultant who writes the plan also build it?
They can, but it must not be a condition, and the findings should be written so that any developer or agency could execute them. A consultant whose recommendations always conclude "hire me to build this" is a salesperson with a nicer invoice. Ask upfront whether the deliverable is hand-able to a third party; if the answer is vague, that is your answer.
Is a software consultant the same as an IT consultant?
Overlapping but not identical. Software consulting is about what to build and how to architect it; IT consulting covers the wider estate — infrastructure, procurement, support models, security and the roadmap across all of it. One person can do both, and on this site the software-shaped work and the infrastructure-shaped work are separate pages because they are separate purchases.