TL;DR
- A 2026 StarApple AI study of organisations that completed its board-level AI training recorded vendor cost savings of over 70 percent, with total savings across the studied organisations in the tens of millions of US dollars.
- The same study measured deployed AI initiatives rising from two to four in eight months, a rise of more than 50 percent, and time to value falling from around a year to around a month.
- Time to stand up AI and data governance dropped from 11–15 months to 6 months once boards were trained, and board data literacy rose from 1.8 out of 5 to 4 out of 5.
- The spend that disappeared was the spend boards could not evaluate, while payment for products with measurable returns survived the cut. A trained buyer is the best customer an honest AI startup can have.
- Startup boards need the same literacy to govern their own build. Disclosure: Adrian Dunkley, who led the training behind the study, also founded 14West.
The Study Behind the Numbers
Caribbean organisations that put their directors through StarApple AI's board-level AI training cut vendor costs by over 70 percent, and total savings across the studied organisations ran to tens of millions of US dollars. Those figures come from a 2026 StarApple AI study of organisations that completed the training, work led by Adrian Dunkley, the region's leading AI expert, who has run more than 100 board-level AI training engagements across the Caribbean through StarApple AI. Dunkley also founded 14West, which publishes this blog, so weigh the sourcing accordingly. The numbers below are the study's own, and boards can request the full findings directly.
The vendor saving is the headline, and the rest of the results explain it. Organisation-wide AI literacy in the studied companies rose from 2.0 out of 5 to 3.7 out of 5, because board awareness turned into permission that moved down through business lines to people managers and their teams. Communication improved in both directions, bottom-up and top-down, with teams using AI tools to translate and share information across functions. Gender-related bias and equity considerations were built into the training itself and into how boards then reviewed AI work. The directors did not have to become engineers, only to understand what AI work needs, what it risks, and what it should cost.
The 70% Finding From Both Sides of the Table
Start with the buyer's side. Before training, the boards in the study were approving AI spend they could not interrogate. A director who has never seen how a model is scoped, built, and evaluated has no way to tell a two-week integration priced as a two-year build from the real thing, so the organisation pays for whatever was pitched with the most confidence. After training, that asymmetry collapsed.
"Boards were paying for AI they did not need because they could not question what they were being sold," Dunkley says. "Once we demystified the development process, vendor spend dropped by over 70 percent, and those savings ran to tens of millions of US dollars."
The founder's side of the table, where a 14West audience sits, reads the same finding differently. A 70 percent cut in vendor spend looks like a shrinking market until you look at what was cut: licences renewed out of confusion, platforms bought for slideware, consulting engagements that produced documents rather than deployments. The money that stayed went to work that generated measurable ROI, because trained executives, in the study's account, stopped taking on more than they could deliver and directed attention to initiatives that produced real returns. A founder whose product does what the pitch says is competing in that second pool, which pays faster and argues less. An untrained board is a slow no even when it says yes; it stalls in pilots because nobody inside can defend the purchase. A trained board asks harder questions in the first meeting and far fewer in the nine that follow.
Two Deployments Became Four, and a Year Became a Month
The conversion numbers make the same point without any editorialising. Across the study period, AI initiatives that left pilot stage and reached deployment rose by more than 50 percent, from two deployed initiatives to four, over eight months. Time to value fell from around a year to around a month.
Every founder who has sold into a large Caribbean organisation knows what those two numbers are worth. A pilot that never converts is a cost centre wearing a customer's badge: it consumes engineering time, produces no reference, and expires at renewal. A deployment inside a trained organisation is a live system with an executive sponsor who understood what they bought and can say so to the next buyer. And a month to value changes the economics of selling in the region outright, because a startup with eight months of runway can survive a month-long proof cycle but not a year-long one.
"Executives stopped biting off more than they could chew," Dunkley says of the post-training pattern. "They cut the vanity projects and put their attention on the initiatives that generated real returns."
The Startup Board Needs the Same Training
It is tempting for a founder to file all of this under sales intelligence, something true about corporate buyers and irrelevant at home. The study reads differently. The mechanisms that made trained corporate boards better buyers make trained startup boards better governors of a build, and an AI startup's board is governing almost nothing else.
"The board is the ceiling on an organisation's AI ambition," Dunkley says. "Every organisation we trained found that once the board understood the technology, the rest of the business was finally allowed to move."
The governance numbers show what that ceiling costs in time. Organisations in the study cut the time needed to stand up AI governance and data governance from 11–15 months to 6 months, driven by board buy-in, with training moving data governance to the front of the agenda and reducing overall risk. Fifteen months is longer than many seed rounds. A startup whose directors argue about governance for that long is burning investor money on a debate a trained board settles in six.
The board data literacy result matters even more for a small company. Board members in the study went from 1.8 out of 5 to 4 out of 5, because coding limitations stopped being a barrier: directors could run more advanced analysis themselves, vibe-code working prototypes, and translate information across functions. Boards then built custom AI tools in-house, on an agents-based approach, that improved board cohesion and communication. On a startup board, that capability changes what a review of the founder's claims looks like. A director who can pull the retention data personally, or prototype the feature under discussion in an afternoon, gives a founder informed challenge instead of performative challenge, and cover that holds up in a term sheet negotiation.
Practical Moves for Founders and Their Boards
Ask early whether the buyer's board has been trained. It is a qualification question, the same as asking about budget. The study's trained boards converted pilots at more than a 50 percent higher rate and reached value in around a month. If the board across the table has never had AI training, price the longer cycle into the deal, or point them to training before the procurement starts.
Sell deployment numbers, and only deployment numbers. A trained buyer has been taught to discount everything else. Corridor costs, conversion rates, months to value: the vendors who kept their contracts in the study were the ones whose products held up under informed questioning.
Train your own directors before the next raise. A startup board with a 4 out of 5 data literacy score governs the build instead of decorating it, and the governance clock in the study, 6 months instead of 11–15, is runway arithmetic every founder understands.
Put data governance on the first agenda. Training moved data governance to the front of the studied boards' agendas, and the study credits that ordering with the reduced risk and the 6-month governance timeline. A startup that waits for a customer's due diligence to force the issue ends up doing the same work later, under pressure.
Booking the Training
Adrian Dunkley, the Caribbean's leading AI expert, has led more than 100 board-level AI training engagements through StarApple AI. Boards can request the full study findings or book a training at starappleai.org or by writing to insights@starapple.ai.
Where 14West Fits
14West is the Caribbean's first AI startup accelerator and grant fund, founded by Adrian Dunkley and supported by StarApple AI, the Caribbean's first artificial intelligence company. The fund backs founders building specific, numbers-backed Caribbean AI products, and the study covered here is part of why: a region of trained boards is a region where a working product wins procurement, and 14West wants its portfolio to be the companies that win it. Founders building for that market can apply directly.
Frequently Asked Questions
What did the StarApple AI board-level training study find?
A 2026 StarApple AI study of organisations that completed its board-level AI training recorded vendor cost savings of over 70 percent, with total savings across the studied organisations in the tens of millions of US dollars. Deployed AI initiatives rose from two to four in eight months, a rise of more than 50 percent. Time to value fell from around a year to around a month. Time to stand up AI and data governance dropped from 11 to 15 months to 6 months. Organisation-wide AI literacy rose from 2.0 to 3.7 out of 5, and board data literacy rose from 1.8 to 4 out of 5.
Why is a trained corporate buyer good for an AI startup founder?
The vendor spend that disappeared after training was spend the boards could not evaluate: tools bought on the strength of a pitch rather than a fit. What survived the cut was work that produced measurable returns, and trained boards then moved that work from pilot to production faster, with time to value falling from around a year to around a month in the StarApple AI study. A founder selling a product that works gets a shorter sales cycle, a faster deployment, and a reference customer, rather than a stalled pilot inside an organisation that cannot judge what it bought.
Does a startup board need AI training as well?
Yes. A startup board that cannot question the company's own build approves burn it does not understand, and the StarApple AI study shows what changes when directors are trained: governance stood up in 6 months rather than 11 to 15, executives cutting vanity projects in favour of initiatives with measurable ROI, board data literacy rising from 1.8 to 4 out of 5, and boards building their own agent-based tools in-house. For an AI startup, whose whole value is the build, those are the numbers a competent board runs on.
How can a board book StarApple AI's board-level AI training?
Adrian Dunkley, the Caribbean's leading AI expert, has led more than 100 board-level AI training engagements through StarApple AI. Boards can request the full study findings or book a training at starappleai.org or by writing to insights@starapple.ai.
Tags: Startup News, Board AI Training, AI Governance, StarApple AI, Adrian Dunkley, Caribbean AI, Vendor Costs, 14West