88% of companies are now using AI in at least one business function — up from 55% just two years ago, one of the fastest technology adoption curves ever recorded. And yet, according to McKinsey, 95% of those same companies have seen zero measurable return on investment. Not low return — zero. Almost everyone is using AI. Almost nobody is winning with it.
Only 27% of businesses using AI describe themselves as confident and in control of their strategy. The other 73% are stressed, overwhelmed, or still waiting to figure it out. This isn't a capability gap — it's a decision gap. Research from Deloitte, PwC, McKinsey, and Writer is consistent: what separates the businesses winning with AI from the ones drowning in the conversation about it comes down to five decisions. Not five tools, not five budgets — five decisions, most of which cost nothing to make.
The 5 Decisions
1 — They owned one problem completely before touching the next one
Businesses failing with AI are trying to improve everything simultaneously. The businesses winning chose one problem — specific, measurable, painful — and didn't move on until it was solved and the results were documented. One win builds the confidence and evidence to tackle the next one. A dozen half-finished experiments build nothing except a larger bill and a more skeptical team.
If the one problem is lead follow-up or new business development, Apollo.io is a clean way to own that single workflow — contact database, sequencing, and lead scoring in one place — before expanding the stack.
2 — They defined what success looked like before buying anything
This sounds obvious. Almost nobody does it. The 27% who are winning share one habit the other 73% skip entirely: they wrote down the metric — a specific, numerical outcome — that would tell them whether an AI investment was working, before they launched it. Not "more efficient." Not "saves time." A number: response time under 4 hours, client onboarding from 6 hours to 90 minutes, follow-up rate from 40% to 95%. Without that anchor, there's no way to know if the tool is working — or to improve it, or justify it to a team.
A tool like Toggl Track can supply the "before" picture — where hours actually go across tasks, clients, and workflows — so there's a real baseline to measure any AI implementation against.
3 — They invested in people and process before they invested in tools
Most companies put 70% of their AI budget into technology and 30% into the people who have to use it. The companies getting real results flip that equation — and the data shows employees at those companies use AI three times more often. The tool isn't the hard part. Getting a team to change how they work is the hard part. The businesses winning explained what problem was being solved, why the tool was chosen, what the expected outcome was, and what success looked like for each person's role specifically. That's not a training program — it's a conversation. It takes an afternoon, not a quarter.
4 — They protected the human time AI freed up, instead of filling it immediately
This is the decision most businesses get wrong in the second act. AI saves five hours a week per person. Within two weeks, those five hours are filled with new tasks, new meetings, new requests — and the team is right back where they started, just with one more tool to manage. The businesses winning made a deliberate choice: when automation freed time, they protected it and assigned it to higher-value work specifically — strategy over administration, relationship building over inbox management, creative output over formatting. The productivity gain only compounds if it's protected.
Reclaim AI is built for exactly this — when AI recovers time in a week, it blocks that time on the calendar for high-value work before meetings can consume it.
5 — They asked for a strategy, not a recommendation, before picking a tool
PwC's 2026 research is direct: companies that crowdsource AI tool decisions — where anyone can propose a new subscription and it gets adopted — end up with impressive adoption numbers and almost no business impact. The businesses winning made someone accountable for the strategy, not just for approving tools — for defining which problems to solve, in what order, with what measurement criteria. The strategy comes before the tool list, every time, without exception. For small businesses, that means a single conversation that answers: what is our biggest bottleneck, what does success look like, and what should we not touch yet? Most businesses skip that conversation entirely and go straight to the demo.
Ask Jordan
We've been using AI for about six months and I honestly couldn't tell you if it's working. We have five subscriptions, nobody hates them, but nobody seems to be noticeably more productive either. I don't know if we're doing something wrong or if we just have the wrong tools.
The tools are probably fine. What I'm hearing is that success was never defined before you started — so there's no way to know whether it's working. That's the most common place this breaks down. Before any of those five tools, what was the single most expensive problem in your business, in time or money?
Which Group Are You In? A Quick Honest Check
Score yourself on the five decisions above. Not where you want to be — where you actually are right now.
- ① Can you name the one problem your AI investment is currently solving?
- ② Do you have a specific number that tells you whether it's working?
- ③ Does your team know why the tool was chosen and what success looks like for their role?
- ④ Is the time your AI freed up actually protected — or did it quietly disappear into more meetings?
- ⑤ Did someone define the strategy before anyone picked a tool?
Five yeses: you're in the 27%. Three or fewer: the same problem as most businesses — and it's fixable, starting with the one you can't answer yet.
The gap between the 27% and the 73% is not capability, budget, or access to better tools. It's the discipline to decide before deploying. The defining advantage now isn't access to AI — every business has that. It's the ability to deploy it with intent: a defined problem, a measured outcome, a team that understands why, time that's actually protected, and a strategy that preceded the shopping cart.
None of those decisions cost money. Every one requires a conversation — which is exactly what a free Jordan session is built for.
Decision five above touches on governance — for the full picture on what's changing in AI data regulation and the governance stack that closes the gap, see the fastest-growing AI risk in small business right now.