Your growth system
Six stages. You are building operating defaults — how much say people have, how much lands on them and when, who decides, what gets communicated, what help is available.
You are past ten people, so this is where we come in. Everything here stays readable — every step and the evidence behind it. From ten people the hard part is no longer knowing what to do. It is that you cannot keep doing it all yourself, and that is where this usually stalls.
{{ nextTitle }}
{{ nextBody }}
One demand reduced, one resource increased, one governance change. Not four.
{{ gateRisk }}
{{ gateQuestion }}
Headcount is not destiny. These are approximate thresholds, not diagnoses.
Agree the mandate
Six minimum conditions. All of them must be true before the rest of this is worth running. If one is not true, the honest move is to fix that first rather than proceed.
{{ mandateNote }}
The climate you are personally creating
Four questions about your own pattern. They are here as a read on the climate, not as coaching.
Senior management's priorities, communication and participation around psychological health preceded later changes in job demands, resources, distress and engagement. Under fifty people, you are that senior management.
Honest note: we are not aware of direct evidence that improving founder wellbeing improves team outcomes. Research supports founder imprinting of employment systems [Moderate · Baron et al. 1996] and senior-management behaviour shaping demands and resources. Treat this as a hypothesis worth testing in your company, not a proven lever.
How growth is changing the way work happens
For each area, write what is true now, the mechanism behind it, and what it costs the business. You do not need all nine. Three honest ones beat nine tidy ones.
"People are exhausted" is a symptom. "We accept urgent work without a capacity check and make up the difference with unpaid availability" is a mechanism. Only mechanisms can be redesigned.
Test the map against what people actually experience
You can still ask everyone. A survey of fifteen people is not anonymous — we never report a group under five, and below about twenty-five we can only report at company level — and it would tell you less than fifteen conversations would. So this stage gives you an interview guide, not a questionnaire.
Interview guide
— When work arrives that was not planned, what happens next?
— Which decisions do you wait on, and how long do you wait?
— What do you own outright, and where is that unclear?
— When was your last uninterrupted block of real work?
— What help is available when a week goes wrong?
— What did you expect when you joined that turned out differently?
Team session outline
1. Name the friction in the work system, not in each other. Forty minutes.
2. Sort each friction into demand, resource, or governance.
3. Agree which three the company would feel most.
4. Say what you will do, and what you will not do yet.
Rules for participation
— Report patterns, not identifiable stories.
— Do not promise anonymity your team size cannot support.
— Close the loop: say what will change, and what will not change yet.
— Never use what people tell you as performance data.
— Route clinical, legal or safeguarding concerns to qualified services, never to a manager.
You have stopped hearing everything directly. That is when measuring starts telling you something you do not already know. From about twenty-five people the Workplace Wellbeing Score runs at this stage, with an expert.
Choose three changes
One demand reduced. One resource increased. One thing about governance. Exactly three — companies fail at this by doing too much.
{{ blockMessage }}
Why should this work?
Five fields per change. Export stays disabled until all of them are complete — a change without a stated reason is an activity, not a design.
{{ logicEmpty }}
What the evidence says about these levers
StrongWorking time. Among organisational interventions, changes to working-time arrangements carry the strongest review-level support for improving the psychosocial work environment and health. [Aust et al. 2023] Worktime control is associated with better work–non-work balance. [Moderate · Nijp et al. 2012]
EmergingThe counterweight. Greater control over working time can reduce the need for recovery, but it can also extend after-hours messaging and overtime when left unbounded. This rests on a single cross-sectional study. Flexibility on its own is not recovery — write the boundary down. [Edvinsson et al. 2022]
EmergingJob control. A work-reorganisation intervention improved mental health, sickness absence and self-rated performance, mediated by job control. This rests on a single study, so treat control and participation as plausible mechanisms rather than proven ones. [Bond & Bunce 2001] A systematic review of interventions to increase employee control reaches broadly supportive but mixed conclusions. [Moderate · Egan et al. 2007]
StrongJob design generally. Motivational, social and contextual work-design features explain substantial variance in job satisfaction, commitment and stress. [Humphrey et al. 2007] Job characteristics predict later wellbeing. [Lesener et al. 2018]
StrongProgram logic. Organisational changes should carry an explicit program logic, be designed with the people affected, and be evaluated on process as well as outcome. [consensus principles · von Thiele Schwarz et al. 2020, the Sigtuna Principles]
Put it into daily work
This stage produces the artifacts — the documents your company keeps and runs. Each one opens prefilled with a worked example you can clear in one click.
The eight artifacts
Before the clock starts
All five, honestly. The most common failure is running this alone.
Six process conditions make organisational changes work: delivery as intended, clear communication, genuine employee participation, visible senior-leadership support, line-manager support, and skilled external facilitation. Content matters too — but content alone, without these conditions, is not enough.
Thirty, sixty, ninety — and the growth gates
Implementation first, always. If the change did not actually happen, nothing the signal shows means anything yet.
Interventions change conditions first — control, support, schedule influence — and health and retention later. Measuring only burnout at six months is weak design. Measure conditions, then process quality, then outcomes, in that order.
Growth gates
Headcount is not destiny. These are approximate thresholds, not diagnoses.
At ninety days
You have run one cycle. Pick your next three — Design restarts, prefilled with what you kept.
Everything you have built
Yours to keep, at any headcount, forever. Each one is prefilled with a worked example — clear it and write your own.
{{ a.name }}
{{ a.blurb }}
Included in every export
Do not expect burnout to fall in ninety days. Structural changes may need longer to reach downstream outcomes, and a short cycle can make people more critical of the work environment before conditions improve. That is expected while conditions are genuinely changing. Conditions move first — interruptions, waiting on decisions, evening work. Wellbeing and retention follow later. Ninety days is one cycle, not the finish.
How this fails: you run it alone, without protected time or follow-through; you ask people what is wrong and change nothing; it becomes a perk while workload and decision systems stay the same; managers are made responsible for wellbeing without authority; flexibility becomes constant availability; the people absorbing overload keep getting recognised for it; it quietly adds meetings; you formalise too early or too late; purpose becomes a claim on people's time.
Thirty, sixty, ninety
{{ reviewIntro }}
{{ r.title }}
{{ r.status }}A change has been marked stuck twice. That is usually the system resisting rather than the change being wrong. That is when having someone in the room helps.
Book a call{{ measureTitle }}
{{ measureBody1 }}
{{ measureBody2 }}
No group under five is ever reported. Between about ten and twenty-five people, reporting is at company level only — team-by-team breakdowns are neither statistically nor ethically defensible at that size.
The evidence, and how we rate it
Work design is how the work is set up. How much say people have. How much lands on them, and when. Who decides. What gets communicated. What help is available. Change these six, and four things tend to move: burnout, sickness absence, whether people stay, and how the work goes.
How we rate evidence
The two-badge rule: practices carry two badges — one for the driver, one for the practice itself. A driver's strong rating never validates a specific implementation.
What we change, and what changes because of it
These are associations and intervention findings, not laws. Effects depend on the setting and on whether the change is actually implemented. Why do this early: employment models founders set at the beginning shaped the structures those companies still carried years later [Moderate · Baron 1996]. That is an argument about persistence, not a guarantee of results.
What the evidence does not support
That every performance practice is good for health. High-involvement management can operate as work intensification, raising pressure without improving the job. [Strong · Wood 2012; Voorde 2012]
That individual coping is the main lever. Training does not change staffing, role ambiguity, scheduling or decision rights. [Strong · Aust 2023; Kjærgaard 2024]
That there is a universal return-on-investment number. Randomised studies are notably less favourable than observational ones. [Strong · Thonon 2023; Backes 2025]
That a survey is an intervention. [Moderate · Tsantila 2024]
Where this comes from
The practice layer draws on Matt Mochary's The Great CEO Within. Every description here is written fresh in our own words. Also worth reading: Andrew Grove's High Output Management, Jim Dethmer and Diana Chapman's The 15 Commitments of Conscious Leadership, and Ken Blanchard's One Minute Manager.
Where we differ. Mochary places individual productivity habits alongside structural ones. The evidence is clearer: individual coping is downstream of work design. So we install structural practices first. The book is also written for venture-scale companies; every practice here is rewritten for a team of five to fifty.
References
Aust, B. et al. (2023). Scandinavian Journal of Work, Environment & Health, 49, 315–329. doi:10.5271/sjweh.4097 · von Thiele Schwarz, U. et al. (2020). The Sigtuna Principles. EJWOP, 30, 415–427. doi:10.1080/1359432X.2020.1803960 · Roodbari, H., Axtell, C., Nielsen, K., & Sorensen, G. (2022). Applied Psychology, 71(3), 1058–1081. doi:10.1111/apps.12346 · Egan, M. et al. (2007). J Epidemiol Community Health, 61, 945–954. doi:10.1136/jech.2006.054965 · Bond, F., & Bunce, D. (2001). J Occup Health Psychol, 6(4), 290–302. doi:10.1037/1076-8998.6.4.290 · Daniels, K. et al. (2017). Ergonomics, 60, 1177–1196. doi:10.1080/00140139.2017.1303085 · Kjærgaard, A. et al. (2024). Am J Ind Med, 68, 250–263. doi:10.1002/ajim.23694 · Tsantila, F. et al. (2024). Int Arch Occup Environ Health, 97, 915–929. doi:10.1007/s00420-024-02096-6
Dollard, M., & Bakker, A. (2010). doi:10.1348/096317909X470690 · Hall, G., Dollard, M., & Coward, J. (2010). Int J Stress Manag, 17(4), 353–383. doi:10.1037/a0021320 · Bakker, A., & Demerouti, E. (2007). J Manag Psychol, 22, 309–328. doi:10.1108/02683940710733115 · Lesener, T., Gusy, B., & Wolter, C. (2018). Work & Stress, 33, 76–103. doi:10.1080/02678373.2018.1529065 · Lesener, T. et al. (2020). Work & Stress, 34, 259–278. doi:10.1080/02678373.2019.1686440 · Humphrey, S. E., Nahrgang, J., & Morgeson, F. (2007). J Appl Psychol, 92(5), 1332–1356. doi:10.1037/0021-9010.92.5.1332 · Leiter, M., & Maslach, C. (1999). J Health Hum Serv Adm, 21(4), 472–489 · Wood, S. et al. (2012). Human Relations, 65, 419–445. doi:10.1177/0018726711432476 · Voorde, K., Paauwe, J., & Van Veldhoven, M. (2012). Int J Manag Rev, 14, 391–407. doi:10.1111/j.1468-2370.2011.00322.x · Nielsen, K. et al. (2017). Work & Stress, 31, 101–120. doi:10.1080/02678373.2017.1304463
Nijp, H. H. et al. (2012). SJWEH, 38(4), 299–313. doi:10.5271/sjweh.3307 · Wendsche, J., Lohmann-Haislah, A., & Wegge, J. (2016). doi:10.18753/2297-8224-75 · Albulescu, P. et al. (2022). PLoS ONE, 17, e0272460. doi:10.1371/journal.pone.0272460 · Edvinsson, J. et al. (2022). IJERPH, 20, 691. doi:10.3390/ijerph20010691
Genedy, M. et al. (2024). J Bus Venturing. doi:10.1016/j.jbusvent.2023.106367 · Bort, J., Wiklund, J., & Yu, W. (2025). Strategic Management Journal. doi:10.1002/smj.3728 · Domurath, A., Taggar, S., & Patzelt, H. (2022). Small Business Economics, 60, 901–927. doi:10.1007/s11187-022-00629-2 · Rangrez, S. N. et al. (2022). Manag Labour Stud, 47, 199–215. doi:10.1177/0258042X221074757 · Ahmadi, E. et al. (2024). BMC Public Health, 24. doi:10.1186/s12889-024-19578-4 · Schreibauer, E. C. et al. (2020). IJERPH, 17. doi:10.3390/ijerph17207446 · Wagner, A. et al. (2022). BMC Public Health, 22. doi:10.1186/s12889-022-12700-4 · Engels, M. et al. (2024). BMC Public Health, 24. doi:10.1186/s12889-024-19102-8 · Pavlista, V., Angerer, P., & Diebig, M. (2021). BMC Public Health, 21. doi:10.1186/s12889-021-11416-1 · Baron, J. N., Burton, D., & Hannan, M. (1996). doi:10.1093/icc/5.2.239 · Harlin, U., & Berglund, M. (2020). doi:10.1007/s11187-020-00383-3
Thonon, F. et al. (2023). Eur J Public Health, 33, 612–618. doi:10.1093/eurpub/ckad092 · Backes, J. et al. (2025). SJWEH, 52, 79–97. doi:10.5271/sjweh.4265 · Graversen, B. K. et al. (2025). SJWEH, 51, 472–482. doi:10.5271/sjweh.4244 · Sorensen, G. et al. (2018). JOEM, 60, 430–439. doi:10.1097/JOM.0000000000001286 · Punnett, L. et al. (2020). Ann Work Expo Health, 64, 223–235. doi:10.1093/annweh/wxaa003