3 key problems with startup hiring data (and how to fix them)

George Morriss
Sep 5, 2018 · 5 min read

Updated January 2019.

I’ve worked in a number of startup and scale-up companies- typically my remit has encompassed growing talent functions looking to scale headcount by upwards of 100% year-on-year. That’s a fairly typical pace for Talentful’s clients.

Many of the leaders I’ve worked with are really keen to either kickstart or fortify the ‘peripheral’ bits that go alongside the actual delivery of great hires into the business. The top two priorities are normally; building a more effective attraction (brand) strategy and wanting to make the team more data-driven. Fantastic priorities, because they show leadership ‘gets’ forward-thinking recruitment strategy- it’s about more than transactional hiring. Those things aren’t peripheral though- they are absolutely central to running a successful, growing company.

That paradigm of thinking is the most essential precursor to scaling a great business- and talent brand and data are usually the two things companies struggle most to understand fully.

Today I’m going to cover a few of the commonalities that most constrain the latter- being data-driven- in startups and scale-ups. I’m also going to propose a couple of easy fixes to help do data ‘properly’. It’s by no means a complete guide (seriously)- but hopefully it’s useful to those just setting out on building a data-driven talent function.

  1. Using incorrect terminology

Ostensibly senior people in the talent space still deem ‘reports’, ‘metrics’ and ‘statistics’ to be synonymous with one another. They’re different things. For clarity’s sake, reports are the cleaned, formatted expression of your data, whereas metrics are standards- or benchmarks- of that data. Statistics are facts obtained through analysis of the data. For example, ‘of 100 previous applicants, 25% felt their interviewer was able to get an accurate sense of their strengths’. That’s a statistic. 📊

This is a problem because other data-driven functions- and leaders of those functions, your stakeholders- understand these things to be different. It undermines your credibility and the respect afforded to you to not be speaking the same language, or not be providing what you said you would. Familiarise yourself with the difference between these things.

2. Jumping the gun on conclusions.

The well-established idiom that correlation ≠causation is too frequently disregarded in favour of delivering ‘actionable insights’ and immediately ‘validating’ the time/energy spent on the data collection exercise. Oftentimes, conclusions by Talent teams are drawn and presented before they’ve collected enough data. This is lazy and really damaging to a business beyond the Talent function. Why? Because you cannot make purposeful progress towards achieving business goals with ill-derived insights, in any department. Default to conservatism before you hit a statistically significant sized data set, from which you can inform decisions.

The utility of data informing progress towards targets is that it allows you to be methodical in your journey forwards, rather than stumbling into success- or failure. Jumping the gun on drawing conclusions is something scale-up Talent functions do all the time. Making the transition to being data-driven is a trend rooted in a need to embolden long-term, measurable, strategic capacity. Thus, data shouldn’t be framed as a quick-fix and leaders should be patient with their desire to inform decisions based on data collected. Be honest and open with senior stakeholders- “this data and the patterns that emerge will inform our decision-making in the long-term”. In the short term, set that expectation clearly and change one variable at a time to find causation.

3. Measuring the wrong- and too many- things.

I’m a huge advocate for not over-engineering what a scale-up Talent function measures. Anchor what you measure in business outcomes. So many teams report on too many recruiting outputs; pipeline volume, number of candidates per stage, number of screens. This creates noise and it doesn’t give you anything actionable that aligns with the outcomes desired by a best-in-class operation.

I’m a vocal proponent of using 3 key data reports- which are outcome-oriented- for measuring performance and informing long-term talent strategy. This is lean (just 3!?) and it’s fit for purpose as they address the business outcomes we’re looking for, rather than arbitrary outputs.

  • Candidate satisfaction (NPS) score, we want happy candidates whether they’re hired or not. This is vitally important for a fledgling employer brand and to inform process relative to market. Reach out to me if you’re interested in implementing a top-class candidate satisfaction strategy.
  • Time to fill, we want to hire quickly. Data on time to fill helps inform process kinks and blockers. I don’t confuse this with time to hire. Time to hire is a component piece of how happy a candidate will be, but this is more oriented towards whether we’re hitting the pace of delivery we need to internally and thus includes time from a requisition being a business ‘need’. We’re essentially adding the time it takes operationally to make a role live to TTH- and minimising the time it takes to do this is important for a startup.
  • Number of hires, fundamentally we need to grow. I look at this monthly on a micro-level, up to the macro at FY-level.

If we keep our candidates happy, and we hire enough of them, quickly- that’s a strategic outcome coming to fruition. It’s much more efficient to spend your energy and time pushing that, than how many screens your team are conducting. Larger teams- go for it. Spare time and energy comes at a premium in the businesses Talentful work with, so focus on what matters.

The prospect-> candidate-> employee journey doesn’t stop with the recruiting process and the true effectiveness of a great organisation’s people strategy (Talent, HR, L+D as a collective) is retention. You should measure that too, as an end-to-end measure of the employee journey. It’s the big one. In fact it’s so big, it needs attention from everyone in the company- it’s everyone’s job to ensure this is done well.

I hope this has been somewhat useful and I’m happy to answer questions or hear people’s thoughts. Do you feel like your Talent team understands these aspects of running a data-driven operation? What 2/3 key metrics are you most interested in, if you had to choose? Feel free to connect on LinkedIn.

Director, Startups @wearetalentful. Alum @univofstandrews, @cloudreach @LabGeni_us.

Welcome to a place where words matter. On Medium, smart voices and original ideas take center stage - with no ads in sight. Watch
Follow all the topics you care about, and we’ll deliver the best stories for you to your homepage and inbox. Explore
Get unlimited access to the best stories on Medium — and support writers while you’re at it. Just $5/month. Upgrade