The Decision Guide
Build, Buy, or Hire: How to Decide.
Every AI project in a small business comes down to the same question: do you buy a tool, build something custom, do it yourself, or bring in help? Here's the framework I use — in the order I use it.
Quick Answer
Should a small business build, buy, or hire for AI?
Start with the problem, not the preferred solution. Define the problem and what it costs you, then look for an off-the-shelf tool first — it's the fastest option when it genuinely solves the problem. If the process creates competitive advantage or affects significant revenue, consider a custom build. If you enjoy technology and have time and support, DIY with AI assistance can work. Choose the smallest option that solves the problem properly.
- →Always evaluate off-the-shelf tools first — fastest when they solve the real problem
- →Custom builds make sense when the process is a competitive advantage or touches large revenue
- →DIY works when the owner has time, patience, and a relief valve for roadblocks
- →Don't hire a full-time tech employee just because AI feels important
- →The best solution is the one the business can afford, control, maintain, and actually use
The Starting Point
Start with the problem, not the preferred solution.
Most owners approach this backwards. They see a demo, read an article, or hear a competitor is “using AI” — and start shopping for a solution before they've defined the problem. That's how companies end up with expensive subscriptions nobody uses.
The right sequence is: understand the process, quantify what the problem costs, then consider the three options in this order — off-the-shelf tool, custom build, do-it-yourself with AI assistance. Each has a place. The mistake is skipping straight to the one that sounds most exciting.
One more thing before the options: the size of the investment is not what determines whether it's worthwhile. The value and frequency of the problem determine what the solution is worth. A $200 tool that solves a daily problem beats a $20,000 platform that solves a quarterly one.
Option 1
Off-the-shelf: fastest — when it solves the actual problem.
An existing tool is often the right answer. Someone else has already built it, tested it, and hired the support team. But only buy when it solves the actual problem without creating unnecessary risk or complexity.
Before signing up, look past the demo and evaluate the things that matter six months from now:
Monthly and long-term cost
Not just the sticker price — per-user fees, usage tiers, and what happens to the price as you grow.
Control and adaptability
Can you shape the workflow to your process, or must your process bend to the tool?
Support quality
When it breaks during your busiest week, who answers — and how fast?
Integrations
Does it connect to the systems you already run, or does it become another island?
Data ownership and export
Who owns the data? Can you export it in a usable format if you leave?
Vendor dependency
What happens to your business if the vendor raises prices, gets acquired, or disappears?
A relatively inexpensive tool can become costly if employees need to change their entire process, data becomes trapped, or the company has to purchase several additional systems to make it useful. Price is one line item. Dependency is the real cost.
I reject a tool when it gives the vendor too much control over a process the business cannot afford to lose.
— Steffen deGraaf, BotLogix
Option 2
Custom build: when the process is the advantage.
A custom system makes sense when the process creates a competitive advantage, affects a large amount of revenue, or cannot be handled properly by an existing tool. If the way you quote, schedule, or serve customers is part of why customers choose you, that process deserves more than a generic tool.
Here's what surprises owners: a custom build doesn't mean you become a software company. The owner focuses on the outcome — explaining the business process, reviewing decisions, testing the result — without personally solving every technical roadblock. That's what the developer or consulting partner is for.
What you get in return is control. Control over the workflow, the data, the user experience, future changes, connections to other systems, and the long-term cost structure. No vendor deciding to sunset the feature your business runs on.
A practical starting point for custom AI implementation is approximately $1,000 per consulting or development day. A problem that can be understood, solved, tested, and implemented in a day may be a $1,000 project. A connected workflow runs higher. The budget question isn't “how much AI can I afford” — it's “what is this problem currently costing me, and what would it be worth to solve it?”
Option 3
DIY with AI assistance: easy to start, harder to finish.
Building with AI assistance can be a great option for an owner who enjoys technology and has enough time to learn. The tools genuinely have made the first 50 percent easier. But be honest with yourself about what happens after the exciting beginning.
Who helps when the integration fails, the data is structured incorrectly, usage limits are reached, or the tool works in testing but not in the real business? A YouTube video may bring you 50 or 60 percent of the way there. The remaining 40 or 50 percent is where the real system gets built — and where many projects are abandoned.
What is your time actually worth?
If your billable value is $150 an hour and the project takes 60 hours of evenings, that's $9,000 of owner time — before maintenance.
Is this the best use of that time?
Every hour spent debugging an API is an hour not spent selling, serving customers, or running the company.
Do you have patience for testing and maintenance?
Processes change, platforms change, models change. Someone has to keep the system working after launch day.
Can other employees understand and use the system?
If the whole thing lives in your head and your accounts, the business has a new single point of failure: you.
AI makes beginning very easy. The real question is: what is the relief valve when they hit a roadblock?
— Steffen deGraaf, BotLogix
The Process
Seven steps, in order. Don't skip step one.
1. Define the problem
One specific process, described clearly: the input, the repeated steps, the people involved, the decisions, the expected output, and the exceptions.
2. Quantify the cost
Hours per week, errors, delayed customers, missed opportunities. If you can't put a rough number on it, you're not ready to buy a solution.
3. Look for an existing solution
Search properly. Ask peers. A suitable off-the-shelf tool is the fastest path when it genuinely fits.
4. Evaluate the risks
Platform risk, ownership, privacy, dependency. Who owns the data, can it be exported, and what happens if you stop paying?
5. Decide if it justifies custom
Is the process valuable or unique enough — competitive advantage, large revenue impact, or nothing existing handles it properly?
6. Decide if DIY is realistic
Do you honestly have the time, the interest, and a support option when you hit the wall?
7. Choose the smallest option that solves it properly
Not the most advanced option. Not the most impressive demo. The smallest one that actually fixes the problem.
The best solution is not automatically the most advanced one. It is the solution the business can afford, control, maintain, and actually use.
— Steffen deGraaf, BotLogix
The Hiring Question
Should you hire a tech person? Probably not full-time. Not yet.
When AI starts to feel important, the instinct is to hire. But a full-time technology salary is a serious fixed cost, and most 5–25 person companies don't have enough continuous technical work to justify it. There are four levels of help — match the level to the actual volume of work.
External project support
Bring in a consulting partner for a defined project with a measurable result. You get senior capability for exactly as long as you need it — and no longer. Best for: your first one or two implementations.
Fractional technology leadership
A fractional CTO or AI officer gives you ongoing strategic guidance a few days a month — vendor decisions, architecture, hiring, roadmap — without a full-time salary. Best for: companies running multiple AI projects at once.
Forward-deployed engineer
An engineer embedded in your operations who builds inside the real workflow, alongside your team, and ships working systems rather than recommendations. Best for: businesses ready to build a connected foundation.
Full-time hire
Makes sense only when there is enough continuous, valuable technical work to fill the role — and when the position is economically sensible on its own numbers. Not because AI is in the news.
Do not hire a full-time technology employee simply because AI feels important. Hire one when the volume and value of the work make the position make sense. Until then, external support gives you the capability without betting a salary on a trend.
Go deeper
Using AI in Your Business: The Owner's Guide
The full pillar — costs, failures, frameworks.
The First 90 Days of AI Implementation
What a real rollout looks like, week by week.
How to Hire AI Help Without Getting Burned
Red flags and the questions to ask before signing.
What Is a Fractional AI Officer?
Ongoing AI leadership without the full-time hire.
AI Strategy Day
$1,000 on-site. A real plan you own.
Custom AI Build
When the process is the advantage, build it properly.
30-minute strategy session
For real businesses with real problems.
Not for tire kickers. Not for tech tourists looking for a demo. For business owners who already know something in their operation is broken — and want to know if AI can fix it.
This isn't about replacing people.It's about killing the bad business processes your team hates as much as you do — the manual work, the missed calls, the after-hours admin.
Strategy sessions are free. Strategy Days are $1,000, capped at 4 people per session. Larger teams or multi-location workshops by special arrangement — ask when you book.
Your session is with
Steffen deGraaf
Founder, BotLogix · Burlington, ON