Forrester: Rising AI Costs to Cut B2C AI Projects by a Third in 2027

Forrester has published its 2027 predictions for B2C marketing, customer experience and digital, forecasting that rising artificial intelligence costs will push consumer-facing brands to pull back on roughly a third of their AI initiatives. The research firm's outlook signals a shift from the current wave of broad AI experimentation toward a more selective, cost-disciplined approach to deployment.

The prediction centers on the economics of scaling AI rather than its capability: as usage grows, so does the compute, licensing and integration spend required to keep AI-powered marketing and CX tools running at scale. Forrester anticipates that many B2C organizations will respond by auditing their AI portfolios and shutting down or shelving initiatives that cannot demonstrate clear return.

For leaders running AI and digital transformation programs, this is a signal that the easy phase of AI adoption — pilot everything, measure later — is ending. Budget scrutiny is catching up with enthusiasm, and finance teams are starting to ask the same ROI questions of AI that they ask of any other technology investment. Organizations that cannot tie specific AI use cases to measurable business outcomes will find those use cases vulnerable, regardless of how promising they looked in a demo.

This also reshapes how CX and marketing teams should plan for 2027: fewer, better-justified AI projects rather than a wide portfolio of speculative ones. The firms that come out ahead will likely be those that built cost-per-outcome tracking into their AI initiatives from the start, rather than those scrambling to retrofit justification once the invoices arrive.

— A third of B2C AI initiatives are expected to be scaled back due to rising AI costs, according to Forrester's 2027 predictions.

A prediction about AI budget cuts is really a prediction about AI discipline — and that's overdue. Too many organizations greenlit AI pilots on excitement rather than evidence, and now the bill is arriving before the business case is finished.

The uncomfortable truth here isn't that AI is getting too expensive — it's that most B2C teams never built the measurement muscle to know which AI initiatives were paying for themselves in the first place. The brands that scale back a third of their AI projects in 2027 aren't failing at AI; they're finally doing the accounting they should have done at launch. The real lesson for operators: treat every AI initiative like a service-design hypothesis from day one — define the customer or efficiency outcome it must produce, price its running cost against that outcome, and kill it early if the math doesn't hold. Cutting AI spend isn't retreat; done well, it's the first sign of a maturing operating model.