Build, Buy or Extend: Choosing Software for a 100-Person Company

A 100-person company should buy software for processes that are the same everywhere, such as payroll, accounting and email; extend its existing tools where they fit most of the need; and build custom software only for the processes that make it different from competitors or that no product handles well. Making this decision process by process, rather than choosing one grand system, avoids both expensive rebuilds and years of workarounds.

Why 100 people is a turning point

Below 50 people, a few SaaS tools, Tally and WhatsApp groups usually cope. Around 100, the cracks show:

• Data lives in eight or ten tools that do not talk to each other.
• The same customer or order is typed into several systems.
• Reports take days because someone must stitch spreadsheets together.
• Licence costs have grown quietly with headcount.

The temptation is to replace everything with one big system. That is usually the most expensive and riskiest option.

A simple framework

For each core process, ask two questions:

1. Is this process standard or distinctive? Standard means every company does it much the same way. Distinctive means how you do it is part of why customers choose you.
2. Do our current tools handle it well? Well means little re-keying, few workarounds and reliable data.

| | Current tools work | Current tools struggle | | --- | --- | --- | | Standard process | Keep | Buy a better product | | Distinctive process | Extend | Build |

Buy: standard processes, better products

Payroll, statutory compliance, accounting, email and document storage are rarely worth building. If your current tool is failing, switch to a better established product.

Extend: good tools with gaps

Often a CRM or accounting system covers most of a process but misses one piece: a quotation calculator, a dealer portal, an approval step. Build that piece as a small application and connect it through APIs. You keep what works and fix what does not, at a fraction of the cost of replacement.

Build: distinctive processes

Custom software pays off where your process is your advantage:

• Complex quoting for engineered or configured products.
• Scheduling and job tracking that reflects how your operation really runs.
• Client portals that make you easier to work with than competitors.
• Operational data that nobody else captures, which you can turn into better decisions.

Connect everything

Whatever mix you choose, the pieces must share data. Plan an integration layer from the start: a clear owner for each type of data, such as customers in the CRM and invoices in accounts, and automated sync between systems. This is what turns a patchwork into a platform.

Common mistakes

• Building commodity functions such as accounting or HR payroll from scratch.
• Buying a large suite for one feature, then paying for modules nobody uses.
• Ignoring integration until data mismatches start causing errors.
• Deciding by demo rather than by walking a real order through the software.

A 90-day plan

1. Weeks 1 to 3: list core processes and score each on the grid above.
2. Weeks 4 to 6: pick the one with the highest cost of pain and agree the approach.
3. Weeks 7 to 12: deliver the first change, whether a new product, an extension or a custom module, and measure the time saved.

Then repeat for the next process. Within a year, the patchwork becomes a connected set of systems chosen for good reasons.

Frequently asked questions

Is custom software always more expensive than buying?

Not over time. For distinctive processes used by many people, a custom build can cost less over three years than licences plus workarounds. For standard processes, buying is almost always cheaper.

How do we avoid vendor lock-in?

Own your data and code, prefer tools with good export and API access, and keep integrations documented.

Who should make these decisions?

The founder or COO with department heads, ideally supported by someone who understands both operations and technology.

What if we get it wrong?

Process-by-process decisions limit the damage. A wrong choice affects one area and can be corrected without touching the rest.

Make the right call for each process

Turbo Bytes Consulting helps mid-sized companies decide what to buy, extend and build, then delivers the custom parts. See our custom software development services, or compare the options for sales teams in Custom CRM vs Zoho vs HubSpot.

Book a 30-minute scoping call and bring your list of tools. We will help you sort them.