i

Description

Un petit groupe de touristes en gyropodes sur une place de Lyon, devant des immeubles anciens.

Source

Romainbehar · original page

Licence

CC0 Public domain

Original photo, unretouched. The background is a version of it stylised by the site.

Pro-innovation bias

What is new is presumed better, and those who hesitate to adopt it are presumed behind.

In brief

• What: we assume an innovation is good, that everyone should adopt it, quickly, and without changes.
• Why: a progress narrative, the prestige of the new, and we mostly hear from successes and promoters.
• Detect: arguments like “everyone is doing it,” claimed benefits without measurement, holdouts treated as backward.
• Avoid: compare against the current state using criteria, require evidence, test at small scale, plan the exit.

On this page

1. Definition
2. The mechanism
3. Where you encounter it
   1. In project management
   2. In management
4. Famous cases
   1. Segway, “bigger than the Internet” 2001-2020
   2. Google Glass: from enthusiasm to withdrawal 2012-2015
   3. Hybrid corn in Iowa, a foundational diffusion study 1943
5. Detection signals
6. Avoidance reflexes
7. With AI: detect, remediate, mitigate
   1. Detect
   2. Fix
   3. Mitigate
   4. Prompts to integrate into your practices
   5. How to integrate them
   6. Precautions
8. Limitations
9. Sources and verification
10. See also
    1. Combines with, or reinforces
    2. Related biases
    3. Counterweights and related reflexes

Definition#

Fictitious illustration

The pro-innovation bias is the presumption that an innovation is beneficial in itself, that it should be adopted by all stakeholders, that it should spread quickly, and that it should neither be modified nor rejected. The term is analyzed by sociologist Everett Rogers, whose book Diffusion of Innovations (1962) is the reference on how new ideas spread; he critiques it in later editions.

Rogers notes that diffusion research studied almost only innovations that succeeded, and treated those who did not adopt them as laggards. But rejecting, abandoning, or modifying an innovation can be perfectly rational. So this bias is not simply a taste for novelty, but an asymmetry in the burden of proof: the old must justify its place, while the new is presumed to deserve it.

The mechanism#

• The progress narrative. We spontaneously associate the new with improvement. Promoters of an innovation are enthusiastic and numerous; its critics are quieter, creating an impression of unanimity.
• Survivorship bias. Innovations that succeeded are visible and taught; those that failed disappear from memory (see Survivorship bias).
• The “modern” identity. Adopting early is valued; hesitation is seen as rigidity. This social pressure aligns with the The Bandwagon Effect (Trend Effect).
• The gap between promise and use. The announced benefits are demo benefits; integration, training, migration, and maintenance costs only appear after adoption.
• Interests. Vendors, consultants, and teams driving change have an interest in adoption, and their voice dominates the available information.

Where you encounter it#

In project management#

• Technology choice: a recent framework, language, or database is preferred over ones that work, based on articles and conferences rather than measured needs.
• Architecture: moving to a trendy architecture (microservices, events, “serverless”) when the problem does not call for it, with a team that does not master it.
• Technology monitoring: monitoring reports only what is new, never feedback from those who abandoned it.
• Migration: replacing a suitable tool with a newer one, while underestimating transition costs and loss of know-how.
• Adopting an AI assistant or tool: organization-wide launch without a value-measurement phase, because “everyone is getting into it.”

In management#

• Organizational transformation: a trendy organizational or methodological model is rolled out as self-evident, and reservations are read as resistance to change.
• Judging people: employees who hesitate in front of a new tool are judged on attitude rather than on their reasons, which may be very sound.
• Reporting: innovation gains are announced by those who championed it, without independent measurement.
• Recruitment and communication: highlighting the modernity of the tech stack as an attractiveness argument, rather than the content of the work.

Famous cases#

Detection signals#

• “Everyone is doing it,” “it's the future,” “we can't stay on the sidelines.”
• The file details the benefits of the novelty and stays vague about costs, risks, and exit.
• Critics are labeled as “resistant to change” without examining their arguments.
• The comparison is not against the improved current solution, but against its version from ten years ago.
• Evidence comes from the vendor, conferences, or promotional articles.
• Adoption is planned everywhere all at once, with no trial phase.

Avoidance reflexes#

1. Compare against the current state using criteria, not image. What specific problem are we solving? What does the current situation cost, measured? What additional value does the innovation bring, quantified?
2. Apply the same evidence standard to both sides. If we ask the old to justify its place, we ask the new as well.
3. Look for failures. For each innovation under consideration, find who abandoned it and why; seek critical feedback, not just testimonials.
4. Try small, measure, then decide. A bounded pilot with indicators chosen beforehand is better than a broad rollout.
5. Plan the exit. Reversibility, exit cost, vendor dependency: plan B is a decision criterion.
6. Listen to holdouts. Their objections are free information; ask them what would need to be true for them to adopt.
7. Accept modification and rejection. Adapting the innovation to context, adopting it partially, or rejecting it are legitimate outcomes, not failures.

With AI: detect, remediate, mitigate#

Prompts to integrate into your practices#

1. Premortem of an adoption. Before deciding to adopt a new technology or method.

We are considering adopting [technology / method / tool] for [goal]. Imagine that, in one year, this adoption has failed.
Give the ten most plausible causes of failure, ranked by likelihood, distinguishing: technical, organizational, economic, human.
For each cause, propose an observable early signal and a risk-reduction measure.
Do not invent specific cases or statistics; if you cite an experience, say whether it is verifiable.

Context: [team, current setup, constraints]

2. The current solution versus novelty. To compare using the same criteria.

Compare, on an identical criteria grid, our current solution (improved with reasonable effort) and [innovation]: [list of criteria we defined beforehand].
For each criterion, indicate what is known, what is assumed, and what would need to be measured to decide.
End with: (1) the conditions under which it would be reasonable to stay with the current solution; (2) the conditions under which to adopt; (3) the least costly test that would allow a decision.

How to integrate them#

• Before the decision: prompt 2 provides the grid, which the team revises; prompt 1 serves as a final safeguard.
• During the pilot: the premortem’s early signals become tracked indicators.
• At review: compare measured results with the initial assumptions, including when the outcome is disappointing.

Precautions#

• AI can invent feedback, tools, or numbers: every reference must be verified.
• Do not share confidential data from your organization or vendors in an unauthorized tool.
• The generated grid uses common criteria: add your own (regulation, hosting, skills, exit cost).

Limitations#

• Pro-innovation bias is described mainly by work in the sociology of diffusion, rather than established by controlled experiments like other biases; it overlaps with trend effects, optimism, and social pressure.
• The opposite exists and is no more reasonable: the status quo bias and rejection of what comes from outside (see the The “Not Invented Here” syndrome) cause good innovations to be missed. The point is not to distrust the new, but to judge it using the same criteria.
• The Segway numbers are orders of magnitude reported by the press and books; they vary by source.

Sources and verification#

• Rogers, E. M. (1962), Diffusion of Innovations, Free Press; 3rd edition (1983) for the discussion of pro-innovation bias; 5th edition (2003).
• Ryan, B. & Gross, N. (1943), “The diffusion of hybrid seed corn in two Iowa communities,” Rural Sociology, 8(1).
• Kemper, S. (2003), Code Name Ginger: The Story Behind Segway and Dean Kamen's Quest to Invent a New World, Harvard Business School Press.
• The New Atlantis, “America at 10 mph” — retrospective on Segway and its forecasts.

See also#