How To Change Company Culture with One Behavior: The 4T Model
2026-01-30
Changing a company’s culture can feel like trying to turn a giant ship – slow and daunting. Many firms in Nigeria and globally pour money into workshops, trainings, lengthy new “culture” documents...yet nothing really changes. People attend the seminar on innovation or fairness, nod along, and go back to business as usual.
Research shows that U.S. corporations spend over $164 billion on training and education yearly, yet most fail to see real behavior change or performance improvement from that investment. The traditional approach – inform (tell people the new values) and train (teach skills) – often doesn’t stick.
So is there a better way? Yes. Instead of grand slogans and broad campaigns, start small – change one key behavior at a time and do it scientifically.
This is the essence of what organizational experts James Elfer, Siri Chilazi, and Edward Chang (in partnership with Harvard researchers) call the 4T Model for behavior change. It’s a four-step approach that has been tested in companies like AstraZeneca and Nationwide Building Society with impressive results. By focusing on one high-impact behavior and designing a targeted intervention around it, they achieved measurable shifts in culture – like more diverse hiring and more inclusive teams – without expensive mass training that doesn’t work.
Let’s break down the 4Ts and see how you might apply them in a Nigerian workplace context:
1. Target a Specific Behavior, Decision, or Outcome
First, pick one very specific behavior to change. Not “improve innovation” in general, but maybe “encourage managers to accept ideas from junior staff” – a concrete behavior. Not “we want a culture of inclusion”, but perhaps “increase the proportion of female candidates interviewed for roles”. Pinpointing one behavior is both the easiest and the hardest step, the experts note. Easy because if you brainstorm, you’ll find dozens of behaviors that tie into a cultural goal; hard because you must ruthlessly prioritize one.
How do you choose? Look at your data or biggest pain points. For example, if your goal is a more innovative culture, maybe the key behavior is “employees openly sharing new ideas in meetings”. If inclusion is the goal, maybe “managers giving fair chance to all CVs regardless of gender or ethnicity” is the behavior. Use evidence: perhaps HR data shows a drop-off of female candidates after CV screening – that points to a targeted behavior needed in hiring managers (like blind screening or adjusted criteria).
In one case, a global telecom company wanted to diversify hires. Instead of generically training about “unconscious bias”, they targeted one decision point: how managers select CVs for interview. That’s specific and measurable (who do they invite? any change after intervention?).
For your context, let’s say you run a fintech startup and notice very few customer complaints get escalated to product changes. You might target the behavior “engineers proactively discuss customer feedback in weekly meetings”. See how narrow that is? It’s actionable.
2. Develop a Theory of Change (Identify Barriers & Levers)
Next, analyze why that behavior isn’t happening currently and hypothesize what could drive it. Basically, diagnose the patient before prescribing medicine. Talk to employees, observe workflows, review literature. For example, if managers aren’t interviewing female candidates, why? Perhaps bias (stereotypes), or maybe the way CVs are presented (names that reveal gender trigger bias), or lack of female applicants in the pool. Each root cause suggests a different fix.
This step is a bit like being a detective or a doctor. Dig deep: maybe employees don’t speak up with new ideas because in the past they were ridiculed (psychological safety issue). Or because the process to submit ideas is cumbersome (structural issue). Your theory of change could be “If we make it safe and rewarding to share ideas, more people will do it”.
There’s no one-size-fits-all here. It often requires iterating – talk to those involved, consider what behavioral science says. For instance, research shows making diversity salient can reduce bias in hiring decisions. That might give you an idea: what if right before managers screen CVs, you remind them of the value of diverse teams or show a quick stat about bias?
The telecom company above did exactly that – their theory was that hiring managers aren’t biased out of malice, but habit and lack of awareness. So their intervention was to prompt managers about diversity right when reviewing resumes. They tested this with an experiment.
In Nigeria’s context, let’s say you want managers to delegate more (to improve performance and growth). Theory of change might be: managers hoard tasks because they don’t trust others or because they feel it’s faster if they do it. So maybe the lever is training managers to build trust and a system of accountability that reduces their fear. Or showing them data that teams that delegate perform better. Essentially, figure out the barrier (trust? skill? incentive?) and design your approach to tackle that.
3. Design a Timely Intervention
Timing is everything. A brilliant intervention delivered at the wrong moment can fizzle. The 4T model stresses delivering your change tactic when it matters – at the exact point the behavior can occur.
For example, telling managers in January “please be fair in hiring this year” is too broad and ahead of time. But giving them a bias checklist or a 7-minute reminder video right before they screen resumes for an active job opening is timely. It’s right at the decision point.
Consider the AstraZeneca example from the researchers: They wanted to improve candidate experience in interviews. They identified the first few minutes of an interview as crucial (candidates often are nervous and that can hurt their performance). They asked, what can an interviewer do in the first 5 minutes to put candidates at ease? They tested different scripts. One script normalized nerves (like “I know interviews can be nerve-wracking; it’s okay to take a pause if you need to”). This simple change, delivered exactly when interviews started, made candidates perform better and feel more comfortable – and interviewers rated those candidates more highly as a result. The timing – at interview start – was key.
So, design something that fits seamlessly into the workflow at the key moment. It could be:
A checklist, pop-up, or reminder built into a software system (e.g., a prompt in your recruitment software: “Have you considered at least 2 female candidates?”).
A change in forms or templates (e.g., a performance review form that forces managers to write at least one area of improvement and one strength for each team member – targeting fairness).
Guidelines or nudges delivered at meetings (e.g., at the start of brainstorming sessions, the team lead explicitly states “wild ideas welcome, no criticism for first 15 minutes” to set a safe tone).
The intervention often is small and simple, not a huge program. It might even be temporary to jolt behavior into a new pattern. The goal is to make it easy for people to do the right thing at the right time.
In a Nigerian bank context, suppose you want to reduce unethical shortcuts. A timely intervention could be a short “integrity reminder” that employees see when logging in each day – like a rotating message with a real customer story or a quote from leadership. It appears right as they start work (timely daily nudge to do right).
4. Test and Iterate (Robustly Test Your Effect)
Now the scientist hat truly comes on. You’ve rolled out an intervention – does it work? The 4T model urges robust testing, ideally with randomized controlled trials (RCTs). That means if possible, have a group that gets the intervention and a similar group that doesn’t, and compare outcomes.
This might sound fancy, but it can be done pragmatically. For instance, the telecom company tested their resume screening video by randomly assigning some hiring managers to watch it and others to proceed as usual. Result? Those who watched the 7-min video ended up interviewing 12% more women on average. That’s solid evidence their targeted approach worked (and far better than generic bias training which studies show often has no effect or even backfires).
If RCT isn’t feasible due to scale, at least track key metrics before and after, and perhaps compare with historical or peer data. The idea is to measure the change. Did the behavior frequency actually improve? Did it lead to desired outcome (e.g., more diverse hires, more innovation proposals, etc.)?
Crucially, testing might show your theory was slightly off or the intervention needs tweaking. That’s fine – better to know and adjust. Maybe an intervention works for one department but not another; you investigate why, refine and test again.
An example from the model: A UK building society relaunched their inclusion strategy. They wanted managers to act on it, not just sign off. They hypothesized that framing the strategy aspirationally (“we know we can do more on inclusion”) vs. complacently (“we are inclusive”) would spur action. They tested two messaging framings with different manager groups. Result: those who got the “we can do more” aspirational message were more likely to initiate inclusion conversations with their teams, and their teams later reported higher engagement and psychological safety. Testing validated that one wording outperformed the other.
In a local scenario, say you implement a new morning huddle format to improve team communication. Measure if after a month, team members report better clarity or if project turnaround improved. If not, gather feedback – maybe the huddles are too long, or wrong time – tweak and try again.
The mantra is: treat culture change experiments like experiments. Have clear metrics, monitor, and be willing to adapt. This beats spending millions on a blanket training and hoping for the best with no follow-up data.
Bringing It All Together – Small Changes, Big Impact
This 4T approach flips the usual script. Instead of broad culture change slogans (which often lead to cynicism), it says: identify leverage behaviors – the ones that have ripple effects – and zero in on them. Change those through nudges or process tweaks at exactly the right moments, and prove they changed via data.
Over time, a portfolio of these small changes adds up. Culture shifts as a sum of shifted behaviors. It’s less glamorous than a “culture transformation taskforce”, but it actually works.
A few more real examples to inspire:
Bias Reduction in Hiring: (We discussed) – Target behavior: how CVs are screened. Intervention: 7-min unconscious bias video + prompt right before screening. Outcome: +13% non-local (or non-traditional background) candidates interviewed, +20% more actually hired. A short video did what day-long bias workshops often fail to do.
Performance Check-ins: A company wanted a feedback culture. Target: managers giving feedback regularly, not just annual review. Theory: managers forget or feel awkward. Intervention: a simple “feedback Friday” email prompt to managers with a tip each week. Timely (end of week reflection), bite-sized. Testing via engagement surveys showed feedback frequency went up in teams that got the nudges versus those that didn’t (hypothetical example, but plausible).
AI Adoption: Perhaps relevant now – companies introduce AI tools but employees don’t use them. Instead of generic “embrace AI” memos, pick a specific workflow to infuse AI. E.g., target behavior: customer support reps using AI to draft email responses. Intervention: integrate an AI suggestion box in the email interface (timely: when writing emails) and train on that specific use. Measure: did average response time drop? Are reps actually clicking the AI suggestions? This targeted approach might yield uptake, versus just telling staff “AI is good, please use it” in a townhall (which many ignore).
One key mindset change here is seeing culture change as a series of small experiments rather than one huge change program. It aligns well with agile thinking – iterate on behaviors, learn, and scale what works.
Starting Small in Your Organization
Even if you don’t have a behavioral science team, you can apply 4T principles:
Focus on one issue at a time. Don’t overwhelm folks with 10 new “cultural values” to adopt. Pick the one that will make the biggest positive difference if changed.
Find the “highest leverage” moment or group to pilot your idea. Maybe start with one department as a testbed.
Use existing data (HR metrics, surveys, etc.) to identify pain points and later to verify changes. If you lack data, even simple before/after counts or a quick survey can provide insight.
Engage people in the change – often the ideas for interventions can come from the ground. For instance, ask managers, “What would make it easier for you to give feedback weekly?” Their answers might guide your intervention design (maybe an app or a checklist).
Embrace trial and error. Frame it as “We’re trying this tweak for 2 months to see if it improves X. If it does, great; if not, we adjust.” This is less threatening than a permanent mandate and encourages participation because staff see it as co-creating better ways.
Celebrate measurable wins. When that hiring experiment leads to more diverse hires, share the news. It builds credibility for this approach and momentum to tackle the next behavior. It also shows employees that these aren’t gimmicks – they’re actually moving the needle, which increases buy-in.
Changing culture is tough, but not because people can’t change – often it’s because we try to change too many things at once in fuzzy ways. The 4T model gives a blueprint to be surgical and scientific about it.
Imagine you implement 3-4 successful behavioral interventions a year – over a few years, your workplace will have significantly shifted habits and norms…almost without people realizing “the culture changed”, but it did, through their actions. That’s more effective than a thousand speeches about culture.
So, whether you’re aiming for more ethical behavior, more innovation, or better customer service culture, remember: big changes start with targeting one small behavior. Change that, prove it, and then move to the next. One step (or shall we say, one behavior) at a time.