The challenge we ran into with Pin’s DEI approach wasn’t the obvious one. We expected resistance to removing demographic signals from the matching model. What we didn’t expect was how hard it would be to validate the model was working as intended—once you remove those signals, you also remove the most obvious way to audit for bias. We had to build a separate evaluation framework that tested outcomes without referencing the demographic fields the model never saw.

If we were doing it again, I’d build the evaluation infrastructure before shipping the fairness feature, not after. The other thing I’d do differently is be more upfront with customers about what “zero demographic data” means in practice—some assumed it meant the model couldn’t consider diversity at all, when the opposite was true: pipelines came back around 6x more diverse precisely because the model wasn’t pattern-matching on who had historically been hired.

One unexpected challenge was discovering that a diversity initiative can fail through good intentions alone. Early efforts focused on attracting broader talent, but the strongest candidates often dropped out because the process felt opaque. In security, people assess organizations the same way defenders assess systems. Friction, unclear expectations, and inconsistent communication signal deeper problems about culture and trust.

I corrected that by tightening every candidate touchpoint. Interview steps were explained clearly, timelines became predictable, and feedback loops were disciplined. That improved completion rates and raised the quality of conversations on both sides. If starting again, I would treat candidate experience as part of risk management from day one, because people who value clarity and accountability usually strengthen both engineering culture and customer confidence.

An unexpected challenge surfaced when diversity goals reached warehouse and service teams. Supervisors supported the idea, but scheduling practices undermined retention. Shift assignments favored familiar personalities over newer employees with potential. That pattern looked operational, yet it quietly damaged trust and advancement. I introduced transparent shift bidding and skill based cross training. Employees gained fairer access to hours, certifications, and visible responsibility. Morale improved because opportunity became measurable instead of manager dependent.

If starting again, frontline systems would be audited before announcing initiatives. Representation improves faster when everyday processes stop leaking credibility. Culture changes only after fairness appears in schedules, not slogans.

Abhishek Shah, Founder & CEO @Testlify, an AI based hiring tool. We’re building product that makes hiring fair, and so naturally it means that we need to live the same internally and there was a lesson learned from that experience than anything else.

The thing that surprised me was not resistance, it was measurement. We put up a formal diversity hiring program with all good intentions but zero baseline data, so when it was six months into the initiative we literally could not say if it was working, because we didn’t define success metrics beforehand. Good intentions with no metrics is just activity.

How we fixed it: we took a step back, established a baseline at each stage of the hiring process—Application, Screening, Interview, Offer—and measured targets at each of the stages instead of just at the outcome. This gave us a window into exactly where bias was creeping into our process, and it was nowhere where we assumed it was.

If I had to do things again: I would define metrics before running the initiative. These diversity programs most frequently fail because it’s not lack of commitment but lack of measurement discipline. You cannot improve what you have not defined.

The biggest surprise for me was how much diversity efforts depend on what the leader does, not just what they say. I could talk about inclusion all day long, but if my actions didn’t match, nothing moved.

Early on, I’d say the right things in a team meeting, but then go back to making decisions the same old way, pulling in the same trusted people, relying on the same networks. The team saw that. And it sent a message I didn’t intend to send.

The turning point was when I made myself uncomfortable on purpose. I reached out to people I didn’t know. I asked for input from team members I hadn’t been leaning on. I let different voices shape real decisions, not just sit in on conversations. That’s when the culture started to shift.

If I could start over, I’d hold myself accountable first, before holding anyone else accountable. The leader sets the tone. If you want a truly diverse and welcoming team, you have to be the first one to do the hard work, not just champion it from a distance.

Bottom Line: Real change starts at the top. As a leader, your actions speak louder than your words. If you want diversity to take root, you have to model it yourself actively, consistently, and even when it’s uncomfortable.

An unexpected challenge was that high performers felt silently threatened. The diversity initiative was meant to widen opportunity, but some top contributors interpreted it as a signal that standards might shift or their effort would be discounted. That private fear created subtle pushback, especially in promotion discussions. I handled it by framing inclusion as a precision issue, not a preference issue. We reviewed how talent was evaluated, where subjectivity entered, and which achievements were consistently overlooked. Data turned a sensitive debate into a quality conversation.

If starting again, I would address perceived loss earlier. Change lands better when people understand that stronger fairness improves talent decisions, team confidence, and long term performance.

If people believe a diversity initiative is merely promoting a slogan versus a change to their work practices then they may hesitate to implement the change. In product and software development teams this is very often exhibited in everyday decisions around how discovery sessions are setup (who is included), the influence of individuals providing input into the product roadmap, people reviewing assumptions surrounding UX design, and whether the same networks and interviewing methodologies are used by the hiring panel. The solution is to make the initiative “practical” in its application (e.g., expanding candidate sources, establishing common and consistent interview scorecards, rotating who leads the product meeting, and conducting user research on individuals who do not fit the “default” profile of a customer).

Instead of launching with a larger message describing the initiative connected to vagueness/completeness of the goals associated with that initiative – start smaller and more operationally focused on one individual workflow (hiring or product discovery) and define clearly how you would like to modify the behaviour in that workflow and then determine whether or not the behaviours have been modified through measuring them. Culture will shift at a quicker rate when people have the opportunity to see the progression of improvements in the processes used to make decisions (based on measurable behaviours) that will lead to them understanding how diversity and inclusion values affect their organisational culture, versus just believing the organisation values diversity and inclusion.

One unexpected challenge was discovering how many employees lacked access to traditional mentoring networks, which limited early participation. We addressed this by launching a virtual mentorship platform with flexible online forums and an algorithm that matched mentors and mentees by goals, expertise, and affinity group so people could connect regardless of background. If I were starting again, I would prioritize building the matching algorithm and forum features from day one so disconnected employees could find mentors sooner. That focus would accelerate participation and strengthen the pipeline for diverse leadership.

One challenge that caught me off guard was how quickly a diversity initiative can lose credibility internally when managers see it as something being “added on” instead of built into how hiring actually works.

I’ve seen leadership teams announce ambitious diversity goals while frontline hiring managers continued using the same interview habits, referral networks, and promotion criteria they’d relied on for years. The resistance usually wasn’t loud or intentional. It showed up in delayed decisions, vague interview feedback, or people defaulting to “culture fit” over actual capability.

In one manufacturing recruitment project, we expanded outreach to attract more women and underrepresented candidates for supervisory roles. The applicant pool improved almost immediately, but the hiring outcomes barely changed at first. When we reviewed the interview process, we realized managers were rewarding familiarity more than performance potential. Candidates from different industries or communication styles were being evaluated more harshly, even when their track records were strong.

What changed things was making the process more structured and measurable. We standardized interview scoring, required written feedback tied directly to job criteria, and included multiple decision-makers in final interviews. Once managers saw they were still hiring high performers, most of the skepticism faded.

One thing I would do differently is spend much more time early on explaining the “why” to middle management. “Diversity initiatives fail quietly when managers feel they’re being managed instead of included.” Getting buy-in upfront would have reduced a lot of friction later and made the initiative stronger from the start.

What we would do differently is involve employees earlier in defining the barriers. Leaders often assume they know where inequity shows up, but the reality is more subtle. It can appear in who gets a smoother handoff or who receives patience during mistakes. It also shows in whose ideas are repeated and later credited to someone else.

If we started again, we would hold structured listening sessions before designing the initiative. We would then compare those themes with actual management practices. This matters because perception alone can drift and data alone can miss the cause. The strongest work comes from listening well, validating patterns, and improving the systems people experience each day.

My initial instinct when we decided we needed to implement DEI was to outsource it. I’m a big believer in keeping our business lean and focusing on our core competency. That logic completely backfired in this case, though. Virtually everyone, including those it was meant to help, resented this kind of top-down, one-size-fits-all DEI, and we quickly discontinued it. What has worked for us to boost our minority hiring and retention numbers is actually working with our existing minority staff to meet their specific concerns.

A particularly surprising challenge associated with diversity was that we realized diversity did not mean just the ability to attract people from different backgrounds; rather, it meant that everyone felt comfortable being part of the organization and growing professionally within the company. Our initial approach was mainly based on attracting people from more diverse backgrounds, but at one point, I realized that some of our new colleagues preferred to remain silent in discussions related to operations and did not voice their opinions even if pressured to do so.

We realized that we needed to change our approach to the way meetings and check-ups took place because our operational meeting during an event week in Los Angeles was disrupted by a new dispatcher noticing problems that other employees failed to address due to their excessive involvement in the conversation.

Going back, I would definitely pay attention to fostering the environment where people can express themselves and not just hire them for their diversity.

As a trial lawyer and CEO for over twenty years, I’ve led inclusion efforts as a former Chairman of the Civil Rights Section for the Louisiana Association for Justice. My perspective is shaped by representing thousands of individuals against powerful insurance companies where fair treatment is the primary goal.

An unexpected challenge was navigating “Between the Ages” generational friction, where different eras of legal practice struggled to align on workplace norms and communication styles. We overcame this by focusing on shared technical challenges, such as implementing firm-wide cybersecurity protocols, which unified the team through a common professional standard.

If I were starting again, I would treat diversity initiatives like a legal defense, using early intervention to bake these values into our firm’s founding documents. Proactively addressing cultural dynamics before they become disputes ensures that everyone is treated fairly from the moment they join the firm.

One unexpected challenge we faced when treating leadership access as a diversity initiative was the initial uncertainty among up-and-coming crew members when we gave them full jobsite ownership. We had stopped assuming that technical skill would automatically produce leadership skill and instead asked employees to manage a project from estimate walkthrough to final inspection. That shift created hesitation around client interaction, vendor coordination, and completing documentation without being told exactly what to do.

We overcame this by pairing clear ownership with consistent support and expectations, asking employees to lead with guidance rather than step-by-step handholding. Team leaders also focused on earlier identification and resolution of issues because responsibility made problems visible sooner. As a result, our leaders grew into their roles faster and client satisfaction increased.

If I were to start again, I would introduce staged ownership and clearer coaching checkpoints from day one. That would reduce initial uncertainty while preserving the accountability that drove the improvement.

Coming from a mechanic background into the luxury finance world, I focused on a “diversity of access” initiative to ensure the watch market wasn’t just for traditional high-net-worth insiders. I am well-placed to discuss this because my business model relies on bridging the gap between blue-collar practicality and luxury assets.

An unexpected challenge was the deep-seated skepticism from buyers who felt they didn’t “fit the mold” of a luxury client and feared being gatekept or overcharged. I overcame this by implementing a radical transparency policy where I disclose my exact margins on every piece, such as a Rolex GMT, to ensure the process is fair for everyone regardless of their background.

If I were starting again, I would have launched my educational content hub immediately to prove that the barrier to entry is information, not social status. Leading with the “buy a watch and then sell it” advice would have empowered a broader range of people to treat these pieces as a store of value rather than an intimidating mystery.

One unexpected challenge I faced with a diversity initiative was that people agreed with the goal in public, but interpreted the purpose very differently in practice. Some saw it as a hiring effort. Some saw it as a training effort. Some thought it was mainly about representation. Others thought it was about improving how teams communicated and made decisions. That created quiet friction because everyone supported the idea, but not everyone was working toward the same version of it.

The way I overcame it was by moving the conversation from broad values to specific behaviors. Instead of leaving the initiative at the level of “we want to build a more inclusive culture,” I helped define what that should look like during normal work. For example, how do meetings run? Who gets invited into planning conversations? How are interview panels structured? How do managers give feedback? How are promotions discussed? Where might strong employees be overlooked because they are not the loudest person in the room?

That shift made the initiative more practical. It also reduced defensiveness because the conversation was no longer about labeling people as good or bad. It was about improving systems and habits. People are usually more willing to examine a process than to feel personally accused. Once the work became concrete, we could make real changes, such as tightening interview rubrics, creating clearer promotion criteria, and making sure quieter voices had structured ways to contribute during meetings.

What I would do differently is spend more time defining the initiative before launching it. Early momentum feels exciting, but if the team does not share the same definition of success, the effort can drift. I would start with a narrower goal, fewer slogans, and clearer examples of what would change day to day.

The main lesson is that diversity initiatives lose power when they remain abstract. People may agree with the principle while still defaulting to old patterns because no one has translated the principle into work habits. If I were starting again, I would make the first phase about clarity: what problem are we solving, what behavior needs to change, who owns the change, and how will we know it is working?

A meaningful initiative has to become visible in decisions, not just language. That is where trust is either built or lost.

One unexpected challenge was that our first ERG lost momentum when leadership tried to steer its agenda instead of letting members run it. We fixed this by stepping back and giving the group autonomy to set meeting topics, projects, and goals. That change helped the group thrive, improved team communication and motivation, and later coincided with a 20% increase in customer satisfaction. If I were starting again I would establish autonomy and clear boundaries for ERGs from day one and treat them as part of our ongoing work culture rather than top-down projects.

Having spent over a decade on the tools and building Make Fencing from the ground up, I found the biggest challenge was overcoming the “old-school” mindset when diversifying our leadership. When I brought Tayla on to head our sales, the unexpected hurdle was integrating her communication-heavy approach into a traditionally rugged trade environment where the “one-man-band” mentality usually rules.

I overcame this by making her “no dumb questions” philosophy a core company value, which helped our installers and clients stay on the same page throughout the build. We then proved the value of diverse expertise by combining Isaiah’s specialized welding skills with our standard installs to offer high-tech solutions like automated gate systems.

If I were starting again, I would hire for these specialized, diverse skill sets much earlier in the journey rather than trying to wear every hat myself. Prioritizing experts in Colorbond fencing and custom automation from day one would have helped us move from small residential jobs to large-scale commercial contracts much more efficiently.

The unexpected challenge when we built our clinic around serving women specifically — a single-population practice rather than a general one — was that the team I assembled needed a much wider range of life experience than I’d initially budgeted for, or else our care would only really fit a narrow slice of the women we said we served.

Our patients span thirty-five to seventy. Working mothers, empty-nesters, single founders, women navigating divorce, women in long marriages, women in same-sex partnerships, women whose primary stress is professional, women whose primary stress is caregiving for aging parents. The clinical work is the same. The way each woman needs the clinical work explained, contextualized, and integrated into her actual life varies enormously.

The clinical team I initially hired was excellent technically and demographically narrow. Within six months we’d noticed that certain patient segments were churning at higher rates than others, and the pattern correlated with whose life experience the team didn’t share. The fix wasn’t more training. It was deliberately broadening who was on the team — by life stage, by family configuration, by professional background — so that any given patient had someone on the team who genuinely understood what her week looked like.

What I’d do differently next time: design the team’s demographic composition before opening, not after. Hiring for clinical excellence alone gets you a competent team. Hiring for clinical excellence and lived-experience range gets you a team that actually serves the audience you said you’d serve.

The technical part of healthcare is solvable with training. The relational part requires the team to look like the audience, at least in aggregate.

The unexpected challenge we ran into at ChainClarity when thinking deliberately about team composition: the tension between moving fast on hiring and building the kind of pipeline that produces genuine diversity in candidates.

When you’re a small startup hiring quickly, the path of least resistance is to hire from your immediate network. That produces a team that looks a lot like the founder — in background, education, geography, and perspective. It’s not malicious, it’s just the mechanics of network-based recruiting. The challenge is that by the time you notice the homogeneity, you’ve already set cultural defaults that are harder to change.

What we tried to do differently, and what I’d do more deliberately from the start: separate the sourcing question from the evaluation question. The sourcing question is: where are we looking for candidates, and does that set of sources systematically exclude certain groups? The evaluation question is: are our criteria actually predictive of success, or do they reflect familiarity bias? These are different problems requiring different interventions.

The thing I’d do differently if starting again: make structured interviews mandatory from the first hire, not once the team is large enough to have an “HR process.” When every candidate gets the same questions evaluated against the same rubric, you eliminate a significant source of subjective variation that compounds over many hires.

The honest limitation I’d acknowledge: for a small founding team, there’s a real tension between “hire the best available person fast” and “invest in a fair, structured process.” We didn’t always resolve that tension the right way, and the compounding effects of early hiring decisions are real.

As a family-run electrical business, we’ve always tried to be more intentional about who we bring into the team, not just hiring from the same circles we’ve always known. It wasn’t a big formal “diversity initiative” for us, more just a decision to widen the net and bring in people with different backgrounds and experiences.

One thing that caught me off guard wasn’t resistance from the team, it was more about how differently people adjust to the reality of working in people’s homes and on live job sites. We brought in someone who was very capable technically, but new to both the trade environment and the customer-facing side of residential work. On paper, everything looked fine, but in practice there were a few early situations where communication with homeowners wasn’t quite in line with how we normally operate.

Nothing serious, but enough to realise that we had focused too much on the hiring side and not enough on how we bring people into the way we actually work day to day.

In our business, especially with “mum and dad” customers, it’s not just about doing the electrical work properly. It’s also about how you speak to people in their home, how you explain what you’re doing, and making sure they feel comfortable while work is going on around them. That part of the job is just as important as the technical side.

What helped us turn it around was slowing the onboarding process down a bit and pairing them more closely with one of our senior electricians for longer than usual. Not just to learn the technical work, but to actually watch how he interacts with clients on site, how he explains things simply, and how he handles unexpected changes without creating stress for the homeowner.

That made a noticeable difference over time. Things that felt awkward at the start started to become more natural because they were seeing it done properly every day, not just being told about it.

If I was doing it again, I’d build that into the process from the start. Not just teaching the technical side of the job, but really being clearer upfront about how we expect people to communicate and behave in someone’s home, because in our line of work that’s a big part of what defines whether someone is a good fit.

The main takeaway for me was that bringing in different people is only one part of it. The other part is making sure you’ve actually set them up to succeed in the environment you work in, not just the job description on paper.