Free quote: 479-547-9241 Contact Get a free quote

Invert the 70/30: Spend Your AI Effort on People and Process

The most repeated point on the Fierce Pharma panel wasn’t about a model or a tool. It was that technology is moving faster than organizations can absorb it, and that most companies are spending their effort in the wrong place. One panelist said organizations are upside down on this: roughly 70% of the time goes to thinking about the technology and 30% to the people and process around it. The recommendation was to invert it. (Part four of five from the panel. Yesterday: when content is cheap, review becomes the bottleneck.)

Where AI effort goes: technology versus people and process Three stacked 100 percent bars. Typical today, per a panelist’s estimate: 70 percent technology, 30 percent people and process. The panel’s recommendation: 30 percent technology, 70 percent people and process. The 10-20-70 rule of thumb: 10 percent algorithms, 20 percent technology and data, 70 percent people and processes. Where the AI effort goes TechnologyPeople and process Typical todaypanelist’s estimateRecommendedthe panel’s inversion10-20-70 ruleheuristic Typical today: 70% technology Typical today: 30% people and process Recommended: 30% technology Recommended: 70% people and process 10-20-70 rule: 10% algorithms 10-20-70 rule: 20% technology and data 10-20-70 rule: 70% people and processes 70%30%30%70%10%20%70% algorithmstech + data
The 70/30 split is a panelist’s estimate from experience. The 10-20-70 rule is a heuristic often attributed to BCG, via Trust Insights; no specific report or method is cited.

Is there evidence for the ratio?

The panelist’s figure was an estimate from experience. A related rule of thumb, often called the 10-20-70 rule, says successful AI adoption depends about 10% on algorithms, 20% on technology and data, and 70% on people and processes. A Trust Insights post attributes it to Boston Consulting Group but cites no specific report or method, so treat it as a heuristic rather than a measured result. The two ideas agree on direction.

Measure what technology frees people up to do

When commercial teams evaluate AI, they usually go straight to top-line return on investment. A panelist suggested reframing it: what is the technology freeing people up to do? Are reps more confident, better informed and able to take on complex problems? Those softer outcomes sit beside the quantitative ones.

Investor relations offered a similar lens. A speaker said their head of investor relations told them that investors care less about how much you spend on AI than about the outcome: the efficiencies you gain and how much faster you move.

I see the same pattern in SEO agencies. The ones putting more human judgment into the work, not less, are the ones getting results.

Start change management early and make it inclusive

The panel’s guidance on adoption was concrete:

  1. Tie communication to “what’s in it for me” for each stakeholder group, not to the technology roadmap.
  2. Include people across the organization in defining the future, so the change is co-created rather than delivered.
  3. Bring marketers into the process early, gathering their feedback and input along the way.
  4. Prioritize rigorously. Tie every initiative to a clear objective so people can see how it connects.
  5. Help people who are behind. The panel asked the room who knew a colleague not yet using AI, and called that the place where the change has to happen. (On skills gaps, see my take on the junior AEO problem.)

Why it matters now

Technology capability isn’t the scarce resource. As the review bottleneck shows, review, governance and adoption are where scaling stalls. Teams that invert the 70/30 are the ones whose people can actually use what’s already available. The skill set that grows out of this is what Jeremy MacDonald calls the marketing engineer.

Rolling out AI across a marketing team? Get in touch.

naturalstateai

Leave a Reply

Your email address will not be published. Required fields are marked *