"We just raised £6,000,000 and now we have funding to grow the team by 50 people in 3-6 months!!!"
But how will you hire?
The problem with the hiring process is that people over-simplify it or default to hiring as a numbers game: 'w' applications generate 'x' calls that lead to 'y' interviews that create 'z' hires - but this does not scale! How are you becoming data-informed? How can you constantly be learning and iterating on your hiring process to become more efficient as you grow? Let’s explore some simple steps you can take today for a data driven approach to your hiring...
Before you get started, understand what metrics you want to follow and where the data will be stored: will it be an Applicant Tracking System (ATS) or some other form of tracking tool? And which tool will you use to create dashboards? There are plenty of tools you can use for free, such as Google Data Studio, PowerBI, QlikView and Tableau. Ultimately you need to create a Single Source of Truth (SOOT) that can be trusted to tell you the true state of affairs.
Define metrics for leadership teams
Here is an example of what leaders will want to see:
Define metrics that the talent team will need to improve
You need to create a SOOT because most current tools don’t really allow you to test, learn and interpret the data well enough, as they can't collate it.
The types of things the talent team will want to see are: a live current pipeline per job with ratios, candidate feedback, reasons for candidates dropping from the process, source of candidates and response rates on head-hunting. All of this should be per candidate and should include the amount of time they have spent in each part of the process (and any other metrics you choose), so define what your team needs to do well.
Top tip: organise and clean the data before it goes into the SOOT.
I would recommend that you create the following dashboards:
In 2018, Google ran 'over 654,680 experiments, with trained external Search Raters and live tests, resulting in more than 3,234 improvements to Search'. That's 595,429 search quality tests; 44,155 side-by-side experiments; 15,096 live traffic experiments 3,234 launches! I think you would have to agree that's pretty remarkable. What we can learn from this is not to assume that what we are doing will always work and be current; we need to constantly be learning and applying what we learn.
How are you constantly improving?
Before setting up tests in your hiring process consider the following:
Imagine, for example, this block in the hiring pipeline: once you send a tech test to developers, not many of them complete it.
Here are two possible A/B tests:
1) Test A: keep the current process as you think it filters the candidates who are interested; test B: start sending a personalised video from a founder/CTO to thank them for taking the test to see if this increases completion rate.
2) Test A: instead of asking candidates to run a 3-4 hour test, ask them to complete a GitHub code review; test B: ask candidates to come and work with you for a week. Never stop experimenting!
Top tip: build a backlog of tests you'd like to conduct and then agree the priority based on their potential impact.
There is strong evidence to show that taking a data driven approach to hiring removes a lot of the bias from the process. In my last blog, I wrote about why you shouldn't hire on instinct and here's one way to do this: a structured interview process combined with a test like The Cambridge Code. In layperson’s terms: setting up a predetermined set of questions defined from a list of attributes/skills you require for the person to be able to do the job (not by whether or not you like them). Sticking to set questions may feel a bit alien and unnatural, but in time you'll get over the awkwardness and you'll soon realise that you're making great hires! At each stage in the process, answers are plotted onto a matrix that allows you to compare outcomes numerically. Combining this with a psychometric test will increase efficacy.
If you don't know how to interpret data or understand how the dashboards work, then it's completely useless, so spend time on training and educating users. If you teach people how to run tests and how to respond to blockages in the pipeline, you can empower them to come up with solutions.
Top tip: invest in your people and train, train, train!
If you’re interested in hearing more about recruitment strategies from scale-up leaders who are pioneering in this space, we recently hosted a Recruitment Strategies for Scale-ups webinar series - you can watch the replays here.