Halden Industrial: Culture Integration After Acquiring a Robotics Start-Up

The campaign addresses a problem the data does not show. A 34-versus-71 engagement split, 41% engineer attrition and the loss of both founders are structural and managerial signals, not an awareness gap. Nobody in that unit is failing to know the values. Posters and town halls change none of the drivers, and running them signals that leadership has misdiagnosed the situation - which raises cynicism and makes the next intervention harder.

Diagnose instead:

• Operating-model collision: imposing small-team autonomy's opposite - grades plus forced distribution - removed what made the unit effective. The engineers did not resist 'Halden'; they resisted a performance system incompatible with how the work is done.
• Loss of legitimacy: the co-founders were the unit's decision-makers and translators. Their departure left no credible voice.
• Procedural justice: the acquirer changed the rules unilaterally. Perceived fairness of how change is decided drives engagement more than the content of the change.

Sequence:

1. Resolve the concrete conflicts - which performance system applies to R&D, who owns the technical roadmap, what decision rights the unit keeps.
2. Retain and visibly empower respected technical leaders from the acquired unit.
3. Define behaviours through mixed working groups, then encode them in promotion criteria and reward - culture follows what gets promoted and rewarded.

If those are aligned, a campaign is optional decoration. If they are not, a campaign is actively harmful. Recommend redirecting the campaign budget into retention packages for the top 50 engineers and a decision-rights redesign.

Four problems, in ascending order of damage:

1. The statistical premise. Forced distribution assumes performance is normally distributed and that managers can discriminate finely among individuals. In long-cycle, interdependent engineering, output is genuinely team-level; forcing a curve manufactures false precision.
2. It punishes collaboration. Engineers who help colleagues reduce their own differentiation chances. In a unit whose value comes from shared technical problem-solving, this is directly counterproductive.
3. It suppresses the risk-taking you bought the start-up for. Rating systems that penalize a failed experiment teach people to stop running experiments.
4. The comparison is invalid, and everyone can see it. A five-point system calibrated on a mature industrial business, applied to a unit doing different work on a different cycle, will mechanically produce a low mean rating in the acquired unit. That gets read as failure, which drives exactly the attrition already observed. This is the single largest driver and the easiest to fix.

Proposal:

• For R&D, use a contribution-based system: team objectives plus individual scope and impact, with calibration of expectations rather than of a distribution.
• If the enterprise insists on distribution, exclude R&D from the forced curve for a defined period, or widen the tolerance band and compare within comparable work only - never across the whole company.
• Separate technical and managerial ladders so senior engineers are not forced into people-management to progress.
• Pilot in the acquired unit and publish the results; this converts a values argument into an evidence argument, which is what will move a CEO attached to the existing system.

Distinguish before acting. External fill can be correct - new markets, new capability, rapid growth. Test it with three diagnostics:

1. Internal promotion rate at the level immediately below director. If it is near zero, the pipeline has no throughput.
2. Whether qualified internal candidates applied and were rejected, and why. 'Not ready' is a finding; 'not visible' is a different and more fixable finding.
3. Time to fill for internal moves. If internal moves are also slow, the problem is process, not supply.

Why 60% matters even if rational: it is self-reinforcing. Heavy external hiring at director level tells every senior manager that the path upward runs outside the company, which raises senior attrition, which creates more vacancies, which are filled externally. That loop is likely already running in the legacy business.

Actions:

• Define a director success profile from actual successful directors, not from a generic competency list.
• Build successor slates of two named people for each critical role, and give them stretch assignments with an executive sponsor - not a training programme.
• Require every external director hire to arrive with a named internal successor development plan as a condition of the requisition.
• Attack time-to-fill by pre-building slates before vacancies exist, rather than by compressing approvals after they do.
• Keep external hiring for genuinely new capability, and say so explicitly - a clear 'build here, buy there' rule preserves the internal promise far better than a vague aspiration.

Tie back to the CEO: 'one culture by year-end' is not achievable by campaign, but the promotion and performance decisions made in the next two quarters will set the culture for years. That is the argument to make.