Contents · 7 min read
01 / 05 · The choice
Which AI proposals deserve the CEO's attention?
In short
Put cost plans and growth bets on the same page. Cost plans matter when they make today's work cheaper. The CEO should give first claim on new funds to the sales problem that most limits growth.
At one industrial B2B company, three to five sellers generate forty percent of revenue. Much of their sales skill remains in their heads.
The CEO has several AI proposals. One makes reports faster. Another updates the CRM. A third helps staff find prices.
Those are sound plans. They save time and cut cost. The AI program reports them as wins, and they are.
But the firm also has a sales problem. Too much of its skill sits with a few people. The wider team cannot use that skill fast enough.
For the industrial firm, hours saved are easy to count. Growth is harder because released time does not create a sale by itself. It must help the firm win more work or serve more work.
So the next step is to find the sales limit with the most weight.
02 / 05 · The limit
What is holding revenue back?
In short
Find the one problem that most limits what the company can win or keep. Test what you sell and how you price it. Then test sales wins, customer loss and delivery capacity.
For the industrial firm, use this top-line value map.
What you sell
Ask whether sales know-how could become part of a stronger offer. In this firm, buyers value sound advice on how the product fits their work. The sign to watch is qualified demand for that offer.
How you price
Reopen price with customer facts and deal data. Include the effect on delivery. Track the margin the firm keeps after discounts.
What you win
Ask whether better bid prep could help the firm answer more good bids. Track bid volume and win rate as separate results.
What you keep
Ask whether better follow-up could help more customers stay. Compare retention with a group from before the rollout.
What you can deliver
Ask whether saved expert time can serve more qualified work. Count the added work, not just the hours released.
This map turns a broad AI strategy into a sales choice. Score each place by impact and feasibility, as an AI readiness assessment does with company-specific bets.
For the industrial firm, the map points to one main limit. Its best sales know-how does not reach the whole team. That points to the next move.
03 / 05 · The spread
Put your best sales knowledge into every deal
In short
Capture how the strongest sellers prepare and follow up. Bring that know-how into the wider team's work when they need it. Keep the final call with the seller.
For the industrial firm, the strongest sellers carry years of product and customer know-how. A new seller cannot gain that skill fast enough through normal training.
The firm can start with real sales work. Its best sellers show how they prepare a bid. They show what they check before a call. They also show how they follow up.
AI can bring those patterns into the work of the wider team. A seller gets the right prep before a bid. The same seller gets help when it is time to follow up.
The AI-augmented sales department keeps the boundary clear. The seller owns the judgment. The work around that judgment stays ready.
For the industrial firm, this changes more than speed. More sellers can use the skill that helps the best sellers win. That skill also stays with the firm when one person leaves.
The CEO can now compare that growth bet with the cost plans.
04 / 05 · The return
Why can the revenue bet keep paying?
In short
Both types of bet can pay. Task automation can show a fast return, then level off once the task is lean. Shared sales know-how can keep improving how the firm wins and keeps customers.
For the industrial firm, the cost plans are useful. Reports take less work. CRM updates take less work. Price checks take less work.
The gains are real, and the company should take them. Yet each gain has a limit. Once the task is as lean as it can be, there is little left to save.
The sales bet works in a different way. Better prep can help more sellers compete for good work. Better follow-up can support retention. Each gain builds the team's skill for the next deal.
GRAIL's industrial B2B analysis draws the same line. Task gains can be clear in the first quarter and level off by month six. Gains from shared know-how can show up in win rate and retention.
The same point shapes GRAIL's AI consulting work. Automation reaches a limit when the task is lean. Added skill can lead to further gains because the company can do more than it could before.
Saved time still needs care. It has no sales value until the firm uses it on more qualified work. The choice now needs a clear way to fund and test the better bet.
05 / 05 · The bet
Fund the best bet, then prove it
In short
Rank each bet by sales impact and feasibility. Fund the strongest one against a group from before the rollout. Expand it only after win rate improves without hiding a loss in margin.
For the industrial firm, the first bet is clear. Put the know-how of its best sellers into the work of the wider team. Then test whether that changes what the team wins.
For your own company, five steps:
Name the sales limit
The CEO and sales head state what most holds back growth. Do this before the team lists AI tools. This is where a useful AI strategy starts.
Rank the bets
The sales head and finance lead score each one for impact and feasibility. Use the same test behind an AI readiness assessment.
Set the group
The sales lead records results from before the rollout. Keep revenue apart from margin. Keep deal speed apart from tool use.
Track the first signs
Use win rate as the main test. Watch qualified bid volume and realized margin beside it.
Expand with proof
The CEO expands the bet when win rate beats the prior group. Check that margin did not fall. Keep tool use as a side measure, not the result.
Do not give saved hours a sales value without a plan for them. Name the extra qualified work those hours will support. Then count whether that work was won.
The next move is to make the chosen bet part of daily sales work.
Put the bet to work
The next task is to turn the strongest revenue bet into daily sales work, while keeping win rate and realized margin as the tests. Prove It Works focuses 60 to 90 days on the functions where AI can pay back most. It can put shared sales knowledge into custom agents and redesigned work, then build the internal ownership needed to test the change against prior results.
Sixty to ninety days in the two or three functions where AI pays back most: custom agents, redesigned work and internal champions. Fixed fee, agreed in writing before work starts.
Questions CEOs ask about this
How should we use AI to grow revenue rather than only cut costs?
Point AI at the problem limiting what your company can win or keep. It drives revenue only if it improves sales or delivery capacity. Rank each bet by sales impact and feasibility. The CEO must choose the best growth bet and test win rate.
Where should we look for AI-enabled revenue opportunities?
Start where the company fails to win or keep work. Check what you sell and how you price it. Then check sales wins, customer loss and delivery capacity.
How can AI improve pricing, conversion, retention or share of wallet?
Measure price through margin after discounts. Measure conversion through qualified bid volume and win rate. For retention or share of wallet, name the offer or follow-up change before you fund the AI work.
Which growth opportunities are large enough to deserve CEO attention?
Choose the bets with the highest sales impact and a feasible path to proof. The CEO should own any bet that changes what the company sells or how it wins.
How should revenue bets be compared with automation projects?
Put both on the same list. Measure automation through real cost removed, and measure growth through sales results. Do not count released hours as revenue unless they create more delivery.
What leading indicators show that an AI growth bet is working?
Start with qualified bid volume and win rate. Then watch realized margin and response speed. Keep CRM completeness and retention apart from tool use.
Founder, GRAIL
Johan has spent fifteen years in management consulting and executive leadership, including as Chief Product Officer of a SaaS company. He founded GRAIL to help European leadership teams work out what AI changes at the core of their business. He is the author of The Execution Revolution and the host of the ThinkRoom podcast.
Related reading
Sources
- AI in B2B industrial companies: the sales gap caused by knowledge held by a few top sellers, and the choice between task automation and shared skill. (accessed October 8, 2026)
- What does an AI-augmented sales department look like?: why seller judgment stays with the person while the work around it stays ready. (accessed October 8, 2026)
- AI readiness assessment: ranking company-specific AI bets by impact and feasibility. (accessed October 8, 2026)
- AI consulting in Stockholm: why task gains can level off while added skill can lead to further gains. (accessed October 8, 2026)
- How to build an AI strategy: starting with what changes in the business, then choosing where to act. (accessed October 8, 2026)
Next scheduled review: January 27, 2027