What Did the Saved Hours Pay For? Change Futurist | October 2, 2026
Ask your team what the hours AI saved last quarter paid for. Most marketing leaders can report the hours. Few can say where the hours went, and the team supplies its own answer when the leader has none.
Josh Stephenson at Marketing Week reported World Federation of Advertisers research on September 30 that puts numbers on the problem: nearly every marketing team now uses AI, and the most common way to measure it is time saved. Three other pieces published the same two days explain why that measurement stalls. David Pralong of McKinsey argues that executives copy AI wins from teams that had standardized their work before any model arrived. Gabriella Rosen Kellerman, David Martin, and Julia Dhar of BCG report in Harvard Business Review that the way leaders describe AI predicts adoption results. Charlotte Rogers covered a B2B panel where Anouschka Elliott explained how the easier number wins the budget meeting.
On Tuesday I wrote about where marketing judgment develops once entry-level roles, vendor-built agents, and agency production move outside the team. This edition covers the capacity that stays inside: the hours your people get back, and who decides what those hours buy.
Replace time saved as your lead AI metric
Stephenson's report, AI use among marketers hits 96% as focus turns to effectiveness, draws on a WFA survey of 54 marketers across 46 brands. The sample is small and the respondents run large advertisers, so read it as a view of the top of the market.
Use of generative or agentic AI in marketing reached 96 percent, up from 45 percent in 2023. Teams describing their adoption as advanced grew from 16 percent to 27 percent in a year. Marketers told the WFA their ambitions split evenly between efficiency and effectiveness at 22 percent each, with a third aiming at everything equally.
Then look at what they track. Time saved, cost savings, and productivity gains each appear on 52 percent of KPI lists. Speed to market appears on 45 percent. Thirty percent of respondents have no AI KPIs at all.
Marketing leaders say they want effectiveness and they report efficiency. A team reads its key performance indicators as instructions, and a team measured on hours saved will produce hours saved.
In the AI capability framework I keep goals above the four pillars for this reason. Time saved is what the Augmentation pillar produces when it works. The goal the saved time serves sits one level up, and a person has to set it.
It reminds me of what Renu Upadhyay, SVP and CMO at Omnissa, said when I interviewed her on The Agile Brand podcast: "I think that time to value of the projects that we decide and anchor on quickly is for me a new success metric, especially when you are onboarding a new technology."
Commentary: The WFA numbers describe teams that finished adopting AI before they decided how to judge it. The 30 percent with no KPIs have the easier fix, since they have no efficiency dashboard to unwind.
Decide what the freed capacity pays for before you pick a tool
Pralong published Why AI's easiest wins are misleading CEOs in Fortune on October 1. McKinsey's latest survey has AI in use at 89 percent of organizations, up one point in a year. The share of high performers, companies that attribute at least 5 percent of EBIT to AI, stayed at 6 percent.
His explanation: the wins executives cite most come from contact centers and software teams. Contact centers had queues, tracked outcomes, and decades of interaction history before any model arrived. Software teams had testing and code review. Those teams added AI to a system that already defined a good result and already checked for it.
Most marketing work has neither property. Pralong's own example is a bank that uses AI to read small-business loan documents two days faster while the application still waits on handoffs among sales, credit, compliance, and operations. Swap in a campaign brief and the example holds.
He cites McKinsey research from July: among organizations early in AI adoption, those that redesigned workflows were 5.3 times as likely to report enterprise-level value, 32 percent against 6 percent. His instruction to CEOs is to name the outcome first, give one leader the redesigned process, and decide whether the freed capacity supports growth, better service, or lower cost.
That last decision is the one marketing teams skip. In the framework I am building for the next book, the analysis can run fully automated while a human stays accountable for direction. Choosing between growth, service, and cost is a direction decision. If the CMO leaves it open, the CFO makes it at budget time, and the CFO's default is cost.
Commentary: Pralong, a consultant, concedes that his profession earns its fees removing waste from existing processes. Marketing operations leaders should make the same admission about three years of AI pilots.
Treat your KPI set as part of the story your team hears
Kellerman, Martin, and Dhar published How Leaders Talk About AI Predicts Adoption on September 30. Their BCG research finds that a company's AI narrative is a material predictor of adoption results. Companies whose leaders tell a clear story built on growth report outsized impact. Companies whose leaders tell a fear-based story, or no clear story, do not. The authors recommend that leaders state expectations plainly, emphasize growth, treat ambiguity as a business risk, and audit the full signal the company sends.
A monthly report is part of that signal. Pralong writes that employees watching an AI rollout ask, "Am I training my replacement?" A leader who reports hours saved every month and says nothing about where the hours go has answered yes without intending to.
It reminds me of what Sangeeta Prasad, Chief Marketing Officer at Slalom, said when I interviewed her on The Agile Brand podcast: "Adapting an organization to use AI is much more difficult than just bringing AI in; there are so many people who just are not comfortable using it. Training people and removing the fear is really the big, big win."
The fourth of the Agile Brand principles asks leaders to respect employees and their time. Telling people what their recovered time is for is the minimum version of that.
Commentary: I could read only the published summary of this piece, so treat the details as incomplete. The recommendation to audit the full signal is the useful one for operators, because the dashboard, the all-hands deck, and the hiring plan usually come from three different people.
Borrow the CFO's test for your AI reporting
Rogers reported on October 1 from a webinar on Marketing Week's State of Brand in B2B research, in How to get B2B brand back in the budget. The survey has 58.4 percent of B2B marketers focusing more on brand in the past year while 47.7 percent say brand is still no budget priority. Short-term focus outnumbers long-term focus 29.7 percent to 8.5 percent.
Elliott, the former global head of marketing at Goldman Sachs Asset Management, gave the reason in one line: "Performance is winning on explanation, not on effectiveness." A marketer can put performance spend in a monthly report and defend it in a budget meeting.
AI measurement has the same problem. Someone can count hours saved by Friday. Effectiveness takes a quarter to show up and an argument to attribute. So the countable number goes in the report, and teams plan toward the number they report.
Elliott's fix for brand transfers directly. Put leading indicators through the funnel, run small tests on specific segments, and bring the evidence to the CFO before asking for more. She also described a firm that runs a monthly prompt audit, entering the questions its buyers ask into the main models and tracking how the company appears. The survey has 54 percent of B2B marketers investing in generative engine optimization, and Elliott's view is that consistent positioning across every channel does most of that work. That is the coherence half of Coherence at Velocity, with a model now reading every surface at once.
Commentary: Elliott and her co-panelist Jon White of Flowtech gave B2B marketers a budget tactic that doubles as an AI measurement plan. Small, segment-level tests with a commercial result attached are how you earn the right to stop reporting hours.
Run the numbers on a team you recognize
A $400M B2B software company runs a 60-person marketing team. The operations lead reports that each marketer saves about 150 hours a year with AI tools, 9,000 hours in total. At a loaded cost of $80 an hour, the team reports $720,000 in annual savings. Hours saved is the only AI KPI, and it goes to the CFO monthly.
Nobody assigned the 9,000 hours. The team spent them on output, and monthly asset volume went from 40 to 95. Pipeline held at $48M. In the budget review the CFO divides 9,000 hours by a 2,000-hour year and asks why a team with 4.5 people of spare capacity needs its open requisitions. The team hears about that question within a week.
Now make the decision first. Before the year starts, the CMO commits 5,400 of the hours, 60 percent, to one program: a brand and demand test in mid-market manufacturing, the segment where the company has its lowest share of voice. The remaining 3,600 hours go back to the teams for training. She tells the team and the CFO the same thing in the same week. The AI KPIs become segment pipeline, win rate, and the company's showing in a monthly prompt audit of the 20 questions those buyers ask.
Segment pipeline grows from $6M to $10M. At a 25 percent win rate, the $4M in new pipeline produces $1M in bookings against $432,000 of redirected capacity. The CFO sees a return on a named investment. The team sees where its hours went.
The tools and the 9,000 hours are identical in both versions. The CMO made one decision before the year started and said it to both audiences.
Do these four things before you close the Q4 plan
Name the destination for freed hours in writing. Choose growth, service, or cost for each major workflow, and tell your team and your CFO in the same week.
Pair every efficiency KPI with an outcome KPI. If you track time saved on a workflow, track pipeline, conversion, or time to decision on the same workflow.
Audit the signals your team receives. Read your last three all-hands decks, your AI dashboard, and your open requisitions side by side, and fix whichever one contradicts the others.
Ask your agencies for their AI KPIs at renewal. The WFA has 59 percent of brands adding AI clauses to agency contracts and 45 percent with little or only some insight into agency AI use. Disclosure of use is the first clause. Disclosure of what the agency measures is the second.
I have sold two companies and acquired three. Every acquisition model has a line for the savings from combining two teams, and that line takes an afternoon to calculate. The work that decided how the integration went was choosing, before close, what those savings would fund, and then saying it to both teams. People on either side of a deal assume the worst about a savings number until someone tells them what it is for.
Ask your team what last quarter's saved hours paid for. If you get more than one answer, write the answer down this week and report against it next month.
Featured this cycle:
AI use among marketers hits 96% as focus turns to effectiveness, study says. Josh Stephenson, Marketing Week, September 30, 2026
McKinsey senior partner: Why AI's easiest wins are misleading CEOs. David Pralong, McKinsey, published in Fortune, October 1, 2026
How Leaders Talk About AI Predicts Adoption. Gabriella Rosen Kellerman, David Martin, and Julia Dhar, Harvard Business Review, September 30, 2026
'Focus without funding': How to get B2B brand back in the budget. Charlotte Rogers, Marketing Week, October 1, 2026