How to Build Blameless Culture

Imagine di last time wey something spoil under your watch. Shipment comot wrong, client collect di bad version of file, one number for board deck turn out to be off. Now imagine di moment wey somebody gats decide whether to tell you.

Dat pause na di whole game.

For di time wey e dey take person to choose between to come meet you and to quietly dey hope say di problem go fix itself, your culture dey show im real shape. If dem come to you early, you go fit act while di problem still small. If dem wait, you go find out later, when e don big and hard and don already touch more people. Wetin dem decide for dat pause dey depend almost entirely on one thing: wetin dem expect go happen to dem when dem speak up.

Na wetin blameless culture really be about. No be to lower standards. No be to let anybody off di hook. Na to make am safe to talk "dis one spoil, and my hand dey inside am" early enough reach say di truth still dey useful.

Di hidden price of blame

Blame dey feel like accountability. But usually e no be.

When something spoil and di first instinct for di room na to find di person wey responsible, people dey learn fast, durable lesson: mistake dey dangerous to dey near. So dem dey stop to report di small ones. Dem dey round dem estimates up to look safer. Dem dey go quiet on di details wey for help you understand wetin actually happen. Di engineer John Allspaw, wey dey write about how teams dey handle outages, put am plain: when people dey fear say dem go name dem, blame dem, and shame dem, dem dey start to hide information, di organization dey stop to learn, and nothing dey done to keep di same failure from happening again.

Notice wetin blame dey buy you. Feeling of resolution, and team wey just don become quieter. Di mistake wey trigger di blame no dey often be di expensive one. Di expensive one na di next mistake, di one wey nobody warn you about because dem watch wetin happen to di last person wey warn.

One grim irony dey under dis. Di people wey dey best positioned to catch problems early na di ones wey dey closest to di work, di ones wey dem hands dey on top am. Na exactly those people blame culture dey teach to stay silent. You go end up blind exactly where you most need to see.

Blameless no mean say consequence no dey

Na here leaders dey get nervous, and di nervousness fair. If dem no ever hold anybody responsible, standards no go collapse?

Dem go collapse, na why blameless culture no ever mean dat. Di clearer term, wey dem borrow from aviation and medicine, na just culture: shared, agreed-upon line between honest error and genuine recklessness. Honest mistake, wey careful person wey dey do reasonable work make, dem dey meet am with curiosity. Wetin happen? Wetin dey di setup wey make dis one easy to get wrong? To knowingly cut safety corner, to hide failure, or to repeat di same careless act after dem don show you di risk na different thing, and dem dey treat am differently.

Di distinction dey matter because e dey protect di right behaviour. You no dey talk say nothing matter. You dey talk say to tell di truth about mistake go never be di thing wey go make dem punish you. Di honesty dey safe. Di recklessness no dey safe. Most people fit live inside dat line easily once dem trust say e real.

Errors usually na system wey dey wear person name

Na here di reframe wey dey make blamelessness practical instead of just kind.

Di safety researcher James Reason spend im career dey study how things dey go wrong for hospitals, cockpits, and power plants, and he draw sharp line between two ways of looking at error. Di person approach dey blame di individual at di sharp end, di nurse wey give di wrong dose, di operator wey hit di wrong switch, and e dey respond with discipline and reminders to be more careful. Di system approach dey assume say capable people go sometimes err because na wetin humans dey do, and e dey ask wetin conditions make di error likely and let am slip through.

Im line na one wey worth to keep: we no fit change di human condition, but we fit change di conditions wey people dey work under.

For Reason model, one single mistake almost never dey cause serious failure on im own. Di bad outcome dey happen when several weaker spots for di system line up at once, unclear instruction, missing check, tired person, tool wey dey make di wrong action easy. Di individual error na di last hole wey di problem fall through, no be di reason why all di holes dey there.

For leader, dis dey change di question entirely. "Who do dis one?" dey give you person to point at and system wey still spoil. "Wetin make dis one possible, and wetin make am hard to catch?" dey give you fix wey dey protect di next person too. Di first question dey feel like progress. Di second one actually na progress.

Di loop wey dey keep you stuck

One pattern dey play out for blame cultures so reliably reach say e almost be script. Something spoil. Dem attach name to am. Dem reprimand di person, maybe send am go retraining, and everybody agree to be more careful. Di case close.

Den, weeks or months later, e happen again. Different person, same failure. And di response na di same: find di name, reprimand, retrain, close. Di team start to believe say dem no get luck with people, say dem just keep dey hire careless ones. Wetin actually dey happen na say nobody ever touch di condition wey dey behind di error. Di confusing form, di missing confirmation step, di deadline wey dey force people to skip di check, all of am still dey sit there, dey wait for di next reasonable person to waka into am.

Blame dey end di investigation early, right at di moment wey e dey get useful. "Human error" dey sound like answer, but na really where di real question dey begin. If your team keep dey make di same kind of mistake with different people, dat no be hiring problem. Na di system dey tell you, clearly, where e spoil. Blameless culture na wetin dey let you hear am, because nobody gats defend dem name first.

When di truth safe, e dey show up faster

Di Harvard researcher Amy Edmondson, wey dey study how teams dey learn, find something wey surprising while she dey research hospital units. Di teams wey report di most errors no be di worst teams. For several cases na di better ones. Dem no dey make more mistakes. Dem dey bring out di ones wey dey already happen, because dem leaders don make am safe to do so.

Na di payoff of blameless culture for one finding. Di errors dey exist either way. Di only variable wey you dey control na whether you go hear about dem in time to do something.

Edmondson dey careful about one trap too, and e worth to hold. To treat every failure as equally fine no be di answer either. Some failures dey sloppy and preventable. Some na di unavoidable friction of complex work. And some dey intelligent, di result of smart bet wey no pay off, di kind of failure wey you actually want more of if dem mean your team to try new things. Leader work no be to celebrate all failure or to punish all of am. Na to tell di kinds apart, out loud, so people go learn which risks dey welcome and which carelessness no dey welcome.

How to build am, for ordinary moments

Dem no dey declare blameless culture for meeting. Na for how you dey react for di first ten seconds after bad news dem dey build am, again and again, until people believe you.

• Watch your face when person bring you problem. Di first reaction na di one wey people dey remember and recalibrate to. Flinch, sigh, sharpened tone, any of dis dey teach di room to bring you less next time. Steadiness here dey do real work.
• Ask wetin happen before you ask who. Put di sequence of events fully on di table, wetin dem know, wetin dem assume, how di situation look from inside, before anybody name become di headline. Di story almost always dey turn out more reasonable than how e first sound.
• Thank di person wey tell you. Especially when e cost dem something to do am. You dey reward di exact behaviour wey you most need, and everybody wey dey watch dey take note. Dis na di cheapest, highest-return thing on di list.
• Run di blameless version of di post-mortem. After something spoil, gather di people wey involved and ask wetin dey di system wey make di error easy and hard to catch. Di output na fixed condition, no be named culprit. Write down wetin you go change, no be who you go watch.
• Dey honest about your own misses. When you talk "I make di wrong call on dat, na dis I learn," you dey give everybody permission to be person wey dey make mistake and recover. Leader wey dey hide dem own errors no get standing to ask anybody else to admit dem own.
• Draw di line clearly and stick to am. Make di difference between honest mistake and reckless one explicit, and den actually honour am. Di protection only dey work if people don see am hold up when dem test am.

None of dis dey complicated. All of dem dey hard, because di pull towards blame strong exactly when you stressed, wey na exactly when e matter pass.

Wetin you dey really build

Team wey dey trust you with bad news na team wey you fit lead through almost anything. You go know di problems while dem still small. You go get di unflattering data instead of di flattering version. People go take di smart risks wey dey move di work forward, because dem know say dem no go hold honest failure against dem.

Di alternative dey look calmer on di surface. Fewer problems reported, fewer hard conversations. Na di calm of team wey don decide say e no safe to tell you di truth, and e dey last right up until di day wey di thing wey dem no tell you arrive all at once.

If dis feel like more than culture problem, sometimes na so e be. Persistent fear, dread before work, or team wey dey look like say e brace for punishment fit point to deeper strain, for dem or for you, wey better meeting no go fix. No shame dey to bring help, whether na outside facilitator for di team or therapist for yourself if di weight of holding everything don start to cost you. Steady leadership dey built on steady person, and dat person get right to need support too.

Sources

• Harvard Business Review, Strategies for Learning from Failure (Amy C. Edmondson)
• The BMJ / PubMed Central, Human error: models and management (James Reason)
• Etsy Code as Craft, Blameless PostMortems and a Just Culture (John Allspaw)