AI Consultant vs. DIY vs. a Big Firm: How to Choose

The Three Paths to AI in Your Business

You've decided AI could help your business. Maybe it's automating a tedious workflow, forecasting demand, or routing customer questions. Good. Now comes the actual decision: how do you make it happen?

You have three realistic options. Build it yourself with off-the-shelf tools. Hire a big consultancy. Or work with a focused practitioner like us. None of these is universally "right." Each path trades off cost, speed, risk, and control in different ways. Your job is to pick the trade-offs you can live with.

Let's be honest about what each one actually gets you.

DIY: When It Works (and When It Doesn't)

Off-the-shelf tools—Zapier, Make, ChatGPT, HubSpot's AI features—are cheap, fast, and surprisingly capable. If your workflow is simple, well-documented, and doesn't require custom logic, DIY is often the smartest move. You can automate lead routing, summarize meeting notes, or trigger email sequences without writing a line of code or spending a dollar beyond your subscription.

The wall shows up when your process is actually yours. When it spans three systems that don't talk to each other. When the rules are too specific for a template. When you need it to scale beyond a handful of tasks or adapt to edge cases that pop up every week.

At that point, you're either duct-taping workarounds or spending more time maintaining the automation than you saved in the first place. DIY is a fantastic starting point. It's a lousy place to get stuck.

If you can describe the task in one sentence and it maps cleanly to a pre-built integration, do it yourself. If you're three levels deep in conditional logic and Googling error messages at 9 p.m., it's time to move on.

Big Firms: What You're Actually Paying For

Enterprise consultancies—the household names—bring serious firepower. Multi-year roadmaps. Compliance expertise. Teams that can embed across departments and shepherd a project through procurement, IT, legal, and the C-suite. If you're a hospital system rolling out AI across 40 locations, that infrastructure is worth every penny.

But their economics require six-figure minimums. You're not just paying for the solution; you're paying for account managers, status decks, change management workshops, and the overhead of a firm that needs to justify its brand. The people actually building your solution are often two layers removed from the people you're talking to.

Timelines stretch. Scope creeps. And when the project wraps, you're usually left with something proprietary, hosted on their infrastructure, requiring their support to maintain. You don't own it. You rent it.

If you're a 40-person distributor or a regional law firm, this is overkill. You don't need enterprise governance. You need the problem solved, the solution handed over, and the ability to move on.

The Focused Consultant: Speed, Ownership, and Trade-Offs

This is where a practitioner like Small Problems fits. We build custom solutions—specific to your workflow, your data, your edge cases—faster and cheaper than a big firm. You get working code or a deployed model, full documentation, and ownership of everything we build. No licensing fees. No vendor lock-in. It's yours.

The trade-off? You won't get 24/7 support. You won't get a glossy portfolio with a dozen case studies in your exact vertical. And we won't hold your hand through internal change management or spend three months aligning stakeholders. We assume you know the problem, you've got budget and buy-in, and you're ready to move.

Best fit: you're a growing business with a specific, high-value problem. You want it solved well, you want to own the result, and you'd rather spend $30k once than $8k a year forever.

Worst fit: you're still figuring out what you need, or you need someone to run internal politics for you, or the solution has to plug into a dozen enterprise systems with compliance requirements we can't touch.

We're not for everyone. But if you're the right fit, this path is faster, cheaper, and leaves you in control.

A Simple Framework: Match the Path to Your Situation

Here's the clearest way to think about it.

Use DIY if you can describe the task in a single sentence and it maps to a pre-built tool. "When a lead fills out this form, add them to this email sequence." "Summarize my meeting transcripts and drop them in Notion." Fast, cheap, no regrets.

Use a big firm if you need enterprise governance, multi-department coordination, or regulatory hand-holding. If the project has to survive procurement, legal review, and a committee of VPs, you need the infrastructure a big consultancy provides.

Use a focused consultant if the problem is specific, the budget is real but not infinite (call it $5k–$100k), and you want to own the solution instead of renting it. You know what needs to happen. You just need someone to build it.

That's it. No tricks, no hard sell. Just match your situation to the path that fits.

What This Looks Like in Practice

Let's make it concrete. Say you're a regional distributor with 50 employees. You want to automate inventory forecasting—right now it's a spreadsheet mess that takes your ops manager six hours a week and still misses half the stockouts.

DIY route: You try a forecasting plugin for your ERP or a Zapier integration. It works for your top 20 SKUs but chokes on seasonal variation, can't handle your regional suppliers' lead times, and doesn't account for the fact that your biggest client orders in weird lumpy patterns. You spend two months tweaking it and give up.

Big firm route: They scope a nine-month engagement. Discovery workshops. Data governance. A custom dashboard with beautiful visualizations. It works, eventually—but it cost $180k, lives on their cloud, and requires their team for updates. Oh, and it took 14 months, not nine.

Focused consultant route: We spend two weeks understanding your data, your suppliers, and your edge cases. We build a custom model tailored to your mix of SKUs, deploy it to run weekly, and hand you the code and documentation. Four weeks, $28k, you own it. Your ops manager updates the input data; the system does the rest.

Not every problem is this clean. But when it is, the difference is obvious.

How to Decide (Honestly)

If you're still not sure, ask yourself three questions.

Can I describe the problem clearly? Not the solution—the problem. If you can explain in two minutes what's broken and what success looks like, you're ready for a consultant. If you're still in "we need to explore AI" mode, you're not.

Do I have someone internal to own the solution once it's built? A custom solution needs a human on your side—someone to update inputs, monitor outputs, and flag issues. If you don't have that person (or can't assign them 2–5 hours a month), a consultant can't save you.

Is my budget between $5k and $100k? Below that, DIY or wait. Above that, you're in big-firm territory whether you like it or not. In that range, a focused consultant is the best value you'll find.

If you answered yes to all three, let's talk. If you answered no to any of them, that's fine—but reconsider your path before you commit time or money to the wrong one.

Still Mapping It Out?

Look, choosing the right path matters more than moving fast. If you're still weighing your options—or you just want a second opinion before signing anything—book a free call. We'll tell you honestly whether we're the right fit, or whether you'd be better off with one of the other paths. No pitch, no hard sell. Just clarity.