AI has made marketing teams faster at drafting, summarizing and generating variants. But faster first drafts are not the same as faster campaigns. In many teams the authoring stage has simply moved the queue to the creative director and the weekly approval meeting.
Two 2026 data points show the disconnect. Gartner’s CMO Spend Survey says AI initiatives now account for an average of 15.3% of marketing budgets, yet only 30% of respondents report mature or fully developed AI readiness. BCG’s AI at Work survey found 74% of frontline white-collar employees use AI regularly. The result is a growing pile of options that still wait for the same few people to sign off.
Faster drafts can create more work
When a campaign variants task moves from several days to an afternoon, the output does not automatically ship. Twenty plausible options can still contain repeated ideas, off-brief angles or claims that need checking. The creative director has more to review, not more time, and finished work still waits for Friday’s approval meeting.
In other words, the bottleneck has shifted downstream. Drafting is no longer the constraint; review capacity and sign-off schedules are.
Decide what improvement matters first
Before racing to more automation, name the business outcome. A campaign team might aim to test three genuinely different creative approaches within the existing budget, or to get approved work into market sooner. The second goal only works if review and approval also move sooner.
Then map the current path from brief approval to launch:
- Where do assets actually stop moving, and why?
- Which claims, brand voices or standards must stay protected?
- Who really has final approval, and how many people believe they do?
This often reveals that drafting speed was never the main constraint. Unresolved product details, a brief that shifts midway, or unclear approval rights hurt more.
Move the keys, not just the tool
For campaign variants, start with an approved brief, substantiated claims, brand examples and clear creative directions. Agree who selects the strongest options, what needs checking and what “ready for review” means. If everything still routes to one creative director, ask why.
Give the person producing the variants a genuine decision right: choose the strongest options and explain the reasoning. That builds judgment, especially for junior staff who no longer learn through drafting grunt work. Invite teams to challenge the process too: a senior writer’s pushback may reveal a quality problem rather than resistance.
Make the saved time count
BCG reports that 60% of regular frontline AI users in marketing save at least a workday a week. But 66% of regular users across functions say they get limited or no guidance on what to do with that time. If the saved hours just trigger another assignment, people will quietly stop mentioning them.
Give the saving a purpose: protected testing time, earlier delivery, normal working hours, or development. And measure the entire process — preparation, generation, review, correction and delivery — including tool and token costs, not simply how quickly the first draft appeared.
Start with one campaign
Pick a recurring campaign process and ask two questions: where is AI already helping after checking and correction are included, and what needs to change for that benefit to carry through to launch? Agree on a revised process, assign ownership, and test it on the next campaign. Then the CMO’s question — what did the business gain — can be answered with launch speed, workload and quality, not just draft velocity.
Source: MarTech
