Lead With Science, Argue With Data
Never argue for the sake of arguing — argue with your data and ask to see theirs.
- Difficulty
- Moderate
- Time to result
- ~months to results
- Steps
- 5
- Confidence
- 90%
When 23andMe was attacked with claims that consumers would misunderstand genetic results, harm themselves, or panic, Wojcicki refused to argue rhetorically. Her standard response became 'show me your data' — because she had over 250 publications demonstrating that customers understood their results, contributed to research, and changed their behavior. The framework has two halves: build the evidence base proactively so it exists before the fight, then use it to invert the burden of proof onto the objector, who almost always has only conjecture. Wojcicki frames this as a habit inherited from running a company of scientists. It is slow to build and decisive to deploy.
Origin
In 23andMe's early years, medical meetings produced a stream of dramatic predictions — that consumers would kill themselves over results, that non-paternity discoveries would cause mass abandonment of women. Wojcicki noticed these were pure conjecture, and that a team of scientists could simply generate the data nobody else had bothered to collect.
Core principles
- 01Conjecture loses to evidence, but only if you have already built the evidence base before the fight.
- 02'Show me your data' inverts the burden of proof onto the objector.
- 03Publishing is a persuasion strategy, not just a scientific ritual.
- 04The data must speak to the specific objection, not to your product's virtue in general.
How to run it
- 1
Catalogue the objections verbatim
Write down the actual claims made against you, in the objector's words: 'people can't understand this information', 'this is dangerous'. Vague fears become testable once written precisely.
- 2
Define the settling evidence
For each objection, specify the study, dataset, or publication that would resolve it. Comprehension objections need comprehension studies; safety objections need safety data.
Pro tip Design the evidence to answer the objection exactly — generic proof of product virtue does not settle a specific fear.
Watch out Evidence that only your marketing team believes is not evidence. Publish it.
- 3
Build the base before the fight escalates
Generate the research during the years the objections are merely social, so it exists when they become regulatory. 23andMe spent years accumulating publications while under fire.
Watch out This is the expensive step and the one everyone skips. Without it, the rest of the framework is just a rhetorical trick.
- 4
Lead with your data in the room
Open with the evidence rather than the argument. 'I have over 250 publications that talk about how people can understand it' is a different conversation from 'I think consumers are capable.'
- 5
Ask to see theirs
Invert the burden of proof. Most objections in a scientific field are conjecture dressed as expertise, and the request for data exposes that without hostility.
Pro tip Ask sincerely, not as a gotcha — occasionally they have data, and that is worth knowing.
In the wild
When critics claimed consumers couldn't understand genetic results, Wojcicki's response was to cite over 250 publications showing customers comprehended their results, contributed to research, and changed their lives — then ask the critic for their data.
→ The objection collapsed for lack of evidence, and the accumulated research base ultimately supported the FDA submissions that restored 23andMe's health reports.
At early conferences, one doctor refused to sit at the table with her; others predicted suicides and mass abandonment of women from non-paternity discoveries. Wojcicki noted that all of it was conjecture about what might happen.
→ None of the predicted catastrophes materialized, and the contrast between confident conjecture and eventual data became the company's argumentative posture.
Common mistakes
Reaching for 'show me your data' with none of your own
The move only works because Wojcicki had 250 publications behind her. Deployed empty, it reads as deflection and destroys credibility with the exact audience you need.
Generating evidence only after the regulator arrives
By the time an FDA warning letter lands, you need the data to already exist. Evidence built reactively costs years — 23andMe's took roughly seven.
Arguing for the sake of arguing
Wojcicki explicitly separates debate from evidence. Winning a rhetorical exchange against an incumbent changes nothing about your regulatory position or your customers.
Is it for you?
Best for
Founders and operators facing expert-establishment resistance in medicine, science, finance, or any field where credentialed opinion functions as a veto.
Not ideal for
Taste-driven or aesthetic categories where the objection is preference and no dataset could settle it, or early companies with no budget to generate real evidence.
From the transcript
“We took this approach again as a team of scientists you lead with science and I don't argue just for the sake of argument I…”
“When people come and they would say like oh 23andMe is dangerous people can't understand it I was like well show me your data because…”
“Frankly it was a lot of conjecture about like what was it that was going to happen.”
From the episode
Anne Wojcicki: How 23andMe is Disrupting the Healthcare Industry
Anne Wojcicki