Most "AI for Business" Advice Is Written for Companies That Look Nothing Like Yours

If the advice starts with "align your data strategy," it's not for you.

Open any business publication this year and you'll drown in AI content. McKinsey reports. Harvard Business Review case studies. LinkedIn posts from consultants explaining how a Fortune 500 bank saved $40 million with an LLM pipeline. It all sounds important. It's also useless if you run a 12-person logistics firm in Lekki or a boutique accounting practice in East Legon.

The problem isn't that the advice is wrong. The problem is that it was written for a company with a Chief Data Officer, a security review board, and a procurement process that takes six months. You have none of those things. You have three overworked people, a WhatsApp Business account, and a spreadsheet that everyone edits at the same time and prays.

Here's what nobody selling AI advice will tell you: the size of your company changes almost everything about what you should actually do.

The advice is calibrated for the wrong constraints

A large enterprise's biggest AI problem is coordination. They have the data, the budget, and the people. What they don't have is agreement — between legal, IT, procurement, and 14 department heads who all want their pet project first. So the advice they need is about governance, change management, and "cross-functional alignment."

Your biggest problem is the opposite. You have complete authority to decide what to do by Wednesday. What you don't have is time, a technical team, or the budget to hire a consultant who charges more per month than your senior accountant earns per year.

When you take advice built for the first problem and apply it to the second, you end up doing things that look serious but produce nothing. You write an "AI strategy document." You form a committee. You spend three weeks evaluating enterprise platforms with per-seat pricing designed for 2,000 seats. Six months later you've spent real money and your invoicing process still takes four hours a week.

What the enterprise playbook gets wrong for you

Three specific things.

One: it assumes you need custom builds. Enterprise AI advice is obsessed with fine-tuning models, building internal RAG systems, and vector databases. For a 10-person firm, this is nonsense. You should be gluing together off-the-shelf tools — Zapier, Make, an OpenAI or Claude API key, a Google Sheet — and shipping something that works in a week. Not commissioning a "proof of concept" that lasts a quarter.

Two: it optimises for scale you don't have. A bank cares about processing 800,000 loan applications a month. You care about not losing the 40 leads that came in through Instagram DMs last week. The tooling, the metrics, and the ROI math are completely different. A solution that "only" handles 200 items a day is a disaster for the bank and perfect for you.

Three: it treats AI as a department. Enterprise advice talks about "AI initiatives" and "AI teams." At your size, AI isn't a department. It's a layer that sits on top of the two or three processes that eat the most time. Client intake. Follow-up. Proposal generation. Reporting. Pick one, automate it properly, then pick the next.

What actually works when you're small

The move for a business your size is unglamorous and it works. You look at where your team is spending time on repetitive, rule-based work. You find the process that costs you the most hours per week — usually something like manual data entry from WhatsApp into a CRM, or drafting the same three types of email 40 times a month, or reconciling payment confirmations across three channels.

Then you build one automation. Not a platform. Not a strategy. One automation that saves one person four hours a week. You ship it in ten days. You measure whether it worked. Then you do the next one.

A firm we spoke to in Ikeja was using two junior staff essentially full-time to sort inbound customer messages across WhatsApp, Instagram, and email into their sales pipeline. An AI agent hooked into their inbox and a shared Airtable now does 80% of that sorting. The two staff members are doing actual sales work instead of triage. Total build time: nine days. Total monthly cost: less than what one of them spends on transport.

That's what the AI conversation looks like at your size. Not a transformation. A series of specific, boring wins that compound.

How to filter advice from now on

When you read an AI article or watch a video, ask three questions before you take any of it seriously.

How big is the company in the example? If it's over 200 people, discount it heavily. The constraints don't map.

What's the implementation timeline? If the case study took nine months, it's not for you. Your entire runway on any single project should be measured in weeks.

What's the tooling? If the answer involves a data warehouse, a machine learning platform, or a "center of excellence," close the tab. If the answer involves an API, a spreadsheet, and a workflow tool, keep reading.

The good news is that the tools available to a 10-person firm in 2026 are genuinely powerful. An operator with clear thinking and a willingness to try things can build automations today that would have required a five-person engineering team three years ago. The bad news is that most of the loud advice is still written for the wrong audience, and following it will waste your money.

The bottom line

You don't need an AI strategy. You need one automation, shipped this month, that removes a specific pain from a specific person's week. Then another. Advice that doesn't help you do that — no matter how prestigious the source — isn't advice for you.

Filter accordingly. Then start building.

---

If you want to see how this applies to your business, book a 30-minute audit with LVD Labs. We'll look at your current setup and tell you exactly where AI and automation can make a measurable difference.