The Prediction Challenge: how start-ups can improve their chances of hiring success

The Prediction Challenge: how start-ups can improve their chances of hiring success

Hiring in start-ups is like a detective solving a case: you want to make a great hire (catch the culprit), but you need to be sure that the candidate (suspect) is the right person. Therefore, you have to look for clues to determine whether someone will be a success (guilty).

It’s a challenge of prediction: brilliant detectives work out which proof points increase certainty in advance (motives, lack of alibis) then look for those points to enhance the probability of success. Brilliant hirers need to do the same. Luckily, the Talent Stack is the Watson to your Sherlock Holmes.

Working out the clues from the red herrings

Start-ups know they don’t want to make a bad hire. The pain of getting it wrong has a drag effect which can take a massive toll on a fledgling company. Emotionally, it’s harder to bounce back to where you were before. Once you’ve made one bad hire, you are more wary of even saying yes to someone else. Strategically, the company could have been swimming round in circles, wasting time, sapping energy and getting no closer to the next growth marker. Financially, it costs a multiple of the person’s salary.

Every start-up ‘knows’ that they want someone who fits in, shares their values and will be an asset to the company. A high-impact hire that will deliver on what is needed and propel company growth. But, how do they ‘know’ a candidate will be all these things?

Although there is no wholly accurate predictor – no Colonel Mustard with the lead piping in the billiards room – start-ups can boost their probability of making a great hire by adopting the following system to look for clues.

Ignore the good CV = good person model

Traditionally, companies have favoured two forms of ‘evidence’ to determine whether a candidate would make a good hire: have they done it before and are they a decent person? For start-ups, this doesn’t work.

Firstly, a person’s experience (their CV) is of low relevance when predicting high performance in start-ups. Disruptive companies are, by their nature, horizon-breaking businesses. There is no precedent. Measuring experience is rendered useless as a predictor, because no one has ever done what you need them to do, in the way you are doing it. If they have got close and they are really senior, you probably can’t afford them anyway.

Secondly, taking a candidate out for coffee or lunch to work out if they are a decent sort is fraught with flaws: casual meetings are easy to blag, full of bias and lead to decisions based on instinct rather than accurate data. In short, they are poor predictors of success. Would you find a suspect guilty if they’d merely done something similar and the detective had a hunch?

Use the six-factor model to find the evidence which matters most

The more relevant evidence you look at, the better chance of nailing the right person. You also need to look at multiple dimensions. Measuring two factors is simply limited: if one of the two misses, you only have a 50% prediction of success. Measure six relevant factors, on the other hand, and if one misses (against five ‘positive’ matches) then your chance of success is a whopping 83%. Given the risks associated with hiring, I’d take those odds any day.

This is what we have built at the Talent Stack. Our six-factor model enables you to look at talent from different angles, while providing data and evidence that is more relevant to start-up success than qualifications and experience, such as a candidate’s action-bias and capacity for risk. We want to help as many start-ups customise and implement the model as possible; every detective needs a good sidekick after all.