Choosing business software can feel like a coin flip: build something custom, or buy an off-the-shelf product and hope it fits. It isn't a coin flip. Knowing how to choose between custom and off-the-shelf software comes down to a handful of factors you can actually score — strategic value, workflow uniqueness, integration depth, and the five-year cost of each path.
We at Craftware have spent 20+ years as a Microsoft Partner helping small and mid-market companies — roughly 20 to 500 employees — make this call, and unwind it when it goes wrong. This guide gives you the method, not a pitch: a three-lane framework, a comparison table, six decision questions, a weighted scorecard, a five-year total-cost-of-ownership example with real numbers, and a screenshot-ready decision tree.
First, Know Your Three Lanes — Buy, Configure, or Build
Most guides frame this as build or buy. In practice there are three lanes, and collapsing the middle one is where a lot of SMBs quietly overspend.
- Buy (off-the-shelf / SaaS). Ready-made software you subscribe to — a CRM, an accounting package, a field-service app; also called COTS (commercial off-the-shelf). Fast, cheap, and proven, but you live within the vendor's roadmap and data model.
- Configure (low-code / platform extension). The lane most articles skip. Instead of buying a rigid product or building from scratch, you extend a platform you already own — departmental apps in Power Apps, or Dynamics 365 and Business Central extended with configuration and light custom code. Cheaper than a build; far more flexible than fixed SaaS.
- Build (custom software development). Bespoke code architected around your data, workflows, and existing systems. Highest upfront cost, but no per-seat tax, no ceiling on complexity, and you own the IP outright.
The real question is rarely "custom or off-the-shelf." It's which lane fits this capability — and the answer can differ app by app inside the same company.
Custom vs. Off-the-Shelf Software at a Glance
Here's how the three lanes compare across nine decision dimensions — no column wins every row.
| Dimension | Off-the-shelf (buy) | Low-code (configure) | Custom (build) |
|---|---|---|---|
| Upfront cost | Low | Low–moderate | Moderate–high |
| 5-year TCO | Rises with per-seat growth | Moderate, controllable | Flat after build; best at scale |
| Time-to-value | Days–weeks | Weeks | Weeks–months |
| Customization / fit | Limited to vendor options | Good, within platform limits | Unlimited — built to spec |
| Scalability | Fine until you outgrow the tier | Strong, then platform ceilings | Built to scale with the business |
| Integration depth | Vendor connectors only | Native across the Microsoft stack | Any system, any depth |
| Data & IP ownership | Vendor holds the keys | You own the data; vendor hosts it | You own the code and data outright |
| Maintenance burden | Vendor-managed | Low; you own governance | Requires a maintenance plan |
| Competitive advantage | None — rivals buy the same thing | Moderate | High — a differentiator you own |
Off-the-shelf vs. custom software, pros and cons in short: off-the-shelf trades fit and ownership for speed and a low upfront cost; custom trades a higher upfront cost and self-owned maintenance for exact fit, deep integration, and no per-seat tax. Low-code sits in between.
The 6 Questions Behind Every Build-vs-Buy Software Decision
Run your project through these six questions — the core of any honest set of software build-vs-buy criteria. Answer each and the right lane usually reveals itself.
- Differentiator or commodity? Payroll, email, and accounting are commodities — buy them. If it's how you win customers or run your core operation, lean configure or build. Commodity → buy. Differentiator → build.
- How unique are your workflows? If your process matches how everyone in your industry works, off-the-shelf fits. Bending three tools and a spreadsheet to force a workflow that's genuinely yours is a configure-or-build signal. Standard → buy. Unusual → configure/build.
- What's the 3–5-year TCO of each path? Compare total cost, not sticker price — subscriptions grow with headcount, a build is front-loaded then flat. Stable headcount → buy. Growing seats plus workarounds → configure/build.
- How fast must you be live? A deadline in weeks favors buying or low-code; a strategic system with room to phase can justify custom. Urgent → buy/configure. Strategic → build.
- How deep are your integrations? The average company runs hundreds of separate applications — MuleSoft's Connectivity Benchmark puts the figure near 900, under a third of them integrated — so how well a tool connects to what you already run is rarely minor. Light → buy. Deep, multi-system → configure/build.
- Must you own the data and IP? If losing vendor access would cripple you, or the logic is a trade secret, ownership matters. Comfortable renting → buy. Must own it → build.
Together, these are the clearest signal of when to choose custom software: a core differentiator, unique workflows, deep integrations, and a multi-year horizon that rewards owning the asset.
Score It: A Weighted Build-vs-Buy Scorecard
Six questions give you a gut read; a scorecard turns it into a number you can defend to a CFO. Score each factor from 1 (clearly favors buying) to 5 (clearly favors building), multiply by its weight, and total the results.
| Factor | Weight | Your score (1–5) | Weighted |
|---|---|---|---|
| Strategic value / differentiation | 25% | ___ | ___ |
| Workflow uniqueness | 20% | ___ | ___ |
| Integration depth | 20% | ___ | ___ |
| Scale trajectory (users & data) | 15% | ___ | ___ |
| Data / IP ownership & control | 10% | ___ | ___ |
| 5-year TCO advantage | 10% | ___ | ___ |
| Total | 100% | ___ / 5 |
Read your weighted total against these bands:
- 1.0–2.3 → Buy off-the-shelf — a commodity a proven product already solves.
- 2.4–3.6 → Configure with low-code or a platform extension — real fit, not a ground-up build.
- 3.7–5.0 → Build custom — strategic, unique, deeply integrated, or something you must own.
Strategic value and workflow uniqueness carry nearly half the weight on purpose — a rebuild can't fix those later. Want the fillable version? We'll send it with the free recommendation below.
The Money: A 5-Year TCO Worked Example
"Off-the-shelf is cheaper" is true on day one and often false by year three: SaaS bills per seat and keeps billing, while a build is front-loaded then flat. Here's an illustrative five-year comparison for a 50-user mid-market operations workflow — billing, dispatch, or order management.
| Cost component | Off-the-shelf SaaS | Low-code (configure) | Custom (build) |
|---|---|---|---|
| Upfront (implementation / build) | $15,000 | $40,000 | $110,000 |
| Annual licensing (per-seat) | $36,000 | $12,000 | $0 |
| Annual maintenance / support | included | $8,000 | $18,000 |
| Annual integration & add-ons | $10,000 | $5,000 | included |
| Annual manual-workaround labor | $18,000 | $6,000 | minimal |
| 5-year total | ~$335,000 | ~$195,000 | ~$200,000 |
Two things stand out. First, the well-fit low-code lane is cheapest here — which is why we recommend it when a workflow fits. Second, the crossover: custom starts far higher (about $128,000 after year one versus $79,000 for SaaS), draws level near year two, and is cheaper from year three on as SaaS's per-seat and workaround costs keep stacking. Grow to 80 or 100 seats and it arrives sooner still.
These are round, illustrative figures to show the shape of the decision, not a quote; real numbers swing with seat count, vendor, and integration scope, so model your own headcount and confirm current pricing before committing.
Your Build-vs-Buy Decision Tree
When you want a fast answer, walk the tree. It's built to be screenshot-friendly for a slide or a Slack thread.
START: What kind of capability is this?
1. Is it a commodity? (payroll, email, accounting, generic CRM)
YES -> BUY off-the-shelf.
NO -> go to 2.
2. Are your workflows unique, but data volume and scale are modest?
YES -> CONFIGURE: extend Dynamics 365 / Business Central,
or build the app in Power Apps.
NO -> go to 3.
3. Is it a core differentiator with deep integrations,
high scale, or strict data-ownership needs?
YES -> BUILD custom.
UNSURE -> score it with the scorecard above, or get a partner's read.
Where Low-Code and Power Apps Fit Between Buy and Build
The configure lane is where Microsoft-stack companies capture the most value: you extend tools you already pay for instead of buying another silo — departmental apps in Power Apps, Dynamics 365 and Business Central extended with configuration and targeted code, and Power Automate to connect the systems you keep.
It has real ceilings, and knowing them is how you avoid a rebuild. The most common is delegation: when a query can't be delegated to the data source, Microsoft's documentation notes that Power Apps processes only the first 500 records by default, raisable to a maximum of 2,000 in app settings but no higher. (Delegation behavior evolves, so check Microsoft Learn for current specifics.) Hit that wall on a growing dataset, or outgrow the platform's formula and API limits, and you've reached the point where custom becomes the lower-risk option — not because low-code failed, but because the workload graduated. It's the same boundary we cover in our Power Apps vs. custom development comparison.
De-Risking the Build (and the Buy)
Every lane carries risk; the goal is to name it upfront and design around it.
If you build, the failure modes are scope creep and an unowned maintenance burden. De-risk with an MVP-first release that proves value before expanding, a fixed scope with phased milestones, a named maintenance plan from day one, and partner selection on verifiable references and platform credentials — not the lowest bid.
If you buy, the quiet risks are vendor lock-in and data-exit costs. De-risk with data-portability terms in the contract (export formats and retrieval on exit, in writing), a clear read on switching cost before you standardize, and attention to per-seat escalation so a cheap pilot doesn't become an expensive standard.
Our Human + AI delivery model — senior Microsoft developers accelerated by AI on scaffolding, testing, and boilerplate — targets the build-side risks directly: it compresses the timeline and cost that used to put custom out of reach for an SMB.
A Representative SMB Example: From Spreadsheets to a Custom Build
The following is an illustrative composite of engagements we see with Greater Houston operations teams — not a specific named client, and the figures are representative ranges rather than one company's audited numbers.
A ~60-person specialty distributor ran billing across an off-the-shelf accounting package, two SaaS tools, and a stack of spreadsheets. Each looked cheap alone; together they cost an estimated 25–35 hours a week in manual re-keying and reconciliation, plus recurring invoice errors that strained customer relationships.
Run through the scorecard, they landed near 4.0 — strategic, unique, integration-heavy — so the fix wasn't more SaaS. It was a configure-plus-build path: extend their Microsoft accounting platform, add a custom billing layer, and integrate the pieces so data flowed once. Illustrative year-one outcome: manual handling down roughly 70–80%, invoice errors down sharply, and a five-year cost curve that beat stacking more subscriptions. The right lane stayed invisible until they scored the decision instead of guessing.
Frequently Asked Questions
What's the main difference between custom and off-the-shelf software? Off-the-shelf is a ready-made product you subscribe to and adapt to; custom is built around your workflows and data, and you own it. Off-the-shelf starts cheaper and faster; custom fits exactly and has no per-seat ceiling. Many companies land in the middle by configuring a platform they already own.
How do I calculate the 3–5 year TCO for build vs. buy? Add every recurring cost, not the sticker price. Off-the-shelf: implementation, per-seat subscriptions grown for headcount, add-ons, integration, and manual-workaround labor. Custom: the upfront build plus roughly 15–20% of it per year for maintenance, no per-seat tax. The year the cumulative lines cross is your decision point.
Where does low-code / Power Apps fit between custom and off-the-shelf? It's the configure lane in the middle. Power Apps and extensions to Dynamics 365 or Business Central build a close-fit app faster and cheaper than a ground-up build, while staying more flexible than fixed SaaS. You graduate to custom at delegation, formula, or scale ceilings.
Can I start with off-the-shelf and move to custom later? Yes — a common, reasonable path. Validate the workflow on a SaaS or low-code tool, then rebuild once requirements and volume are proven. Budget for the migration, since data and logic don't transfer automatically, and secure data-export terms upfront.
What happens to my data if my off-the-shelf vendor shuts down or hikes prices? That's the core vendor lock-in risk, and your leverage is contractual: get data-portability and export terms in writing, know your switching cost in advance, and prefer vendors with standard export formats. When data ownership is business-critical, it argues for the configure or build lanes.
The Bottom Line: How to Choose Between Custom and Off-the-Shelf Software
You don't decide this by instinct — you score it. Run the six questions, total the scorecard, and model the five-year TCO, and the right lane stops being a guess: buy the commodities, configure the good-fit workflows, and build what's strategic and yours to own. Getting that call right upfront is what saves you from a rebuild — the most expensive outcome of a build-vs-buy software decision.
Get a Free Build-vs-Buy Recommendation
Not sure which lane your project belongs in? We at Craftware will run your situation through the same scorecard you just read, model a rough five-year TCO, and give you a straight buy / configure / build recommendation — an honest read from a Microsoft Partner that delivers all three, so we have no reason to steer you toward the more expensive path. You'll also get the fillable scorecard to keep. Book your free build-vs-buy recommendation and decide with numbers, not a coin flip.
