Effective Recruiting Metrics for Fast-Growing Startups

We spent the better part of 2015 working with fast-growing tech companies on their internal HR & recruiting analytics. Effective recruiting can be incredibly challenging: First Round Capital’s State of Startups 2016 Report assessed that talent acquisition is founders’ biggest concern for the second year in a row. Recruiting metrics help to identify problems and optimize the talent acquisition process: if you can’t measure it, you can’t improve it. Below we share some of the most common metrics we observed across startups that serve as the basis for understanding and improving your recruiting process. The list below isn’t comprehensive, nor applicable for everyone, but it’s a good place to start.

Many novel applicant tracking systems like Greenhouse and Lever provide certain metrics out-of-the-box, while other metrics are a bit harder to wrangle because they require pulling data from disparate software systems. Unfortunately, pulling these metrics can be different from one company to the next, because each company uses a different cocktail of systems across recruiting/HR, and each company has a different inclination towards recording high quality data on candidates and employees. Likewise, these metrics will differ based on the structure of your recruiting process — number of phone screens, if you use pre-screen software like TripleByte, etc. If you need any guidance on how to derive any of this metrics or discuss how you can benchmark against others (as well as your past self), feel free to reach out via email with any questions.

Source of Hire

What are your most effective recruiting channels? This metric helps you understand where you’re getting the most and the best applicants from — from referrals, college career fairs, your careers website, AngelList, Hired.com, etc. This is useful in understanding if you’re allocating your sourcing resources in optimal ways. If a certain source is underperforming, consider dropping it or diagnosing how to fix it. Consider slicing this metric by job requisition, department, or level of seniority. This analysis can also help identify if there are specific channels that are helpful in finding specific types of candidates: for example, post junior/brand ambassador roles on WayUp because they are geared towards students and recent grads, but save engineering roles for StackOverflow, Hired, etc. It’s also useful to merge this metric with performance and quality, to assess sources of high-quality versus low-quality hires — we discuss Quality of Hire below.

Cost per Hire

How expensive is it for you to hire? Average cost per hire = Total recruiting cost / number of hires made. The total recruiting cost can include various channel fees (e.g. career fairs, advertising, agency fees, SaaS pricing), internal recruiting resources (i.e. recruiter salaries, bonuses, etc.), employee referral bonuses, and time spent interviewing (i.e. number of hours × average salary of interviewer). Consider slicing this by job requisition, department, or level of seniority as well to get a sense of which positions are most expensive to fill.

One thing to note is that on-site interviews can be costly because they take up significant employee time and have second-order impacts on productivity. If this is salient in your cost per hire metric, consider taking a closer look at your funnel metrics and interviewer metrics (both detailed below) to see if you may be able to improve the efficacy of the earlier stages of your interview process.

Quality of Hire

How effective is your team at not only making hires, but also hiring, developing, and retaining high-performing and happy employees? This metric can be a bit elusive but helps establish the overall efficacy & purpose of your recruiting process. Quality can be assessed by merging recruiting data with employee engagement surveys (e.g. from Culture Amp), performance feedback/reviews (e.g. from Reflektive), and employee retention & promotion rates (e.g. from Namely). Consider integrating your ATS with your HRIS, performance management system, and engagement/culture system — this is generally easier said than done. Likewise, often times the data is sparse, missing, or overly qualitative that it proves to be of marginal utility.

The value of Quality of Hire is that it can reveal potential correlations between candidate & interview characteristics and future performance. For example, even if employee referrals are lower cost than agency hires, how do they compare on performance outcomes? Does the number of interviewers a candidate meets with impact performance outcome — so do we need two separate phone screens if we could see the same results with just one?


Last year, Pinterest was one of the first to lead the charge in publicizing their diversity metrics alongside many other startups. To measure diversity, it’s important to track ethnicity and gender ratios. Pinterest slices these by business unit / seniority (e.g. total company, leadership, business, tech, tech interns, etc.). It is important to also look at diversity metrics in the context of employee onboarding, development, and retention. To act upon diversity or look closer at metrics, consider the services of a company like Paradigm that can assist with assessing and combating unconscious & institutional biases.

Candidate Experience / Satisfaction

The job market is competitive and word travels fast. It’s important to treat candidates well, as you may need to revisit their candidacy in the future and they may be a positive (or negative) referral source. Consider surveying candidates after they complete your interview process in order to get effective feedback. If candidates are rejecting your offers or dropping off in your process, this will help diagnose why. For example, you can use Culture Amp for this.

Funnel Analysis

Looking at conversion rates from each stage of your recruiting process to the next can help inform why you’re experiencing drop-off through your process.

Example: [(# of applications) ÷ (# of phone screens)] → [(# of phone screens) ÷ (# of on-sites)] → [(# of onsites) ÷ (# of offers)] → [(# of offers) ÷ (# of hires)]

We can pull apart this funnel to learn various things:

(# of applications) ÷ (# of phone screens)

  • Are there enough applications coming in at the top of the funnel in the first place? If not, we likely have a sourcing problem and need to find more ways to fill the top of the funnel. Try sourcing tools and marketplaces like Hired, Entelo, Connectifier, LinkedIn Recruiter, Whitetruffle, AngelList, etc.
  • Is the job effectively being promoted across channels (e.g. careers page, AngelList, college career fairs, etc.)? Are we searching in the right places for candidates? Consider job board services like ZipRecruiter, Indeed, BroadBean, etc.
  • Is our job description effective, compelling, and reasonable? Consider Textio to optimize. Do our expectations match market availability? For example, are we looking for a Senior Machine Learning Engineer with 10+ years of experience and a PhD to fill a junior level web developer role? Are we being too picky and not interviewing enough candidates, or are we casting too wide of a net and interviewing way too many candidates? At times, there are discrepancies between what functional leaders and managers are looking for, and what frontline recruiters are seeing is likely & available in market.
  • Do we have enough resources to triage applications and sift signal from noise? Consider products like Stella or Atipica.

(# of phone screens) ÷ (# of on-sites)

  • Is our phone screen process effective and structured? Is it too stringent or too lenient? Are we conducting too many phone screens, or are we passing too many candidates through directly to on-site without an effective screen?

(# of offers) ÷ (# of hires)

  • In an ideal world, this number if 1 : 1, but the job market is competitive and the best candidates often have many offers to choose from. Did we effectively sell our company through the interview process? Did we show the candidate love and follow up effectively after making an offer? Did we discuss a timeline with the candidate to create a sense of urgency? Did we make a competitive offer that is inline with market compensation and the candidate’s expectations?

Interviewer Analysis

As aforementioned, having an effective phone or recruiter screen process is essential because on-site interviews can be costly and take up valuable employee time. A phone/recruiter screen can be ineffective in two ways:

  1. The screener passes through too many candidates that are rejected during the on-site or don’t match hiring manager/team expectations;
  2. The screener rejects or doesn’t pass along candidates that are actually qualified and worth interviewing.

You can assess the efficacy of screeners by correlating their decisions with candidate’s on-site interview scorecards, hiring outcomes, or even better, employee performance outcomes. For example, if a phone screener consistently approves candidates that perform very poorly in an on-site, they may not be calibrated to the team’s expectations. Of course, this is dependent on the process by which candidates move from phone screen to on-site, the quality of your interview process, and who decision makers are in the interview stack — i.e. if there is a hiring manager involved, etc. This analysis is of course imperfect because we can’t know the quality or interviewing outcomes of the candidates the screener turns away (because they never made it that far) — we can account for this, in part, via Loss Analysis, mentioned below.

Equally, you can assess the quality of your on-site interviewers in this way. In general, it’s important that every interview have a purpose, and it’s recommended that you have a consistent process / framework by which to conduct interviews and collect feedback. Applicant tracking systems like Greenhouse facilitate this well. There are various services that can standardize or improve the initial screening process — for example, HackerRank, Lytmus, RemoteInterview, CodeFights, etc. — as well as commonly used tools like CoderPad.

Loss Analysis

It’s important to understand which companies you’re competing with for the best-fit candidates. To learn this, you can assess:

  • Where did candidates that we rejected go to work?
  • Where did candidates that rejected our offers go to work?
  • Where did employees that left our company go to work?

For example, if we are a startup and we are rejecting candidates that are going on to work at very admirable competing companies, perhaps we are too harsh in our process or we aren’t assessing the right things. If candidates are rejecting us, perhaps we have a branding problem, are unable to sell candidates on the role, are not paying competitively, etc. One way to do this is through follow-up surveys, as mentioned above. Another way to do this is to diligently look up candidates on LinkedIn after the fact, but there can be a significant lag-time between when the candidate interviews with you and subsequently updates their profile after starting a new job.

Time to Hire

By analyzing the time an applicant spends in each stage of your recruiting process, you can identify bottlenecks that are impacting your hiring rates. Because the recruiting process is so competitive, it’s important to move quickly — if you don’t, someone else will, and a great candidate will be off the market in no time. Important ratios and the questions they answer:

  • Days from application to phone screen: Do you have enough frontline resources to review applications? Are recruiters / hiring managers getting in front of candidates quickly after they apply?
  • Days from phone screen to on-site: Do you have an efficient means to find schedule availability for team members? Do we have a recruiting coordinator? Consider a product like GoodTime or Reschedge, or leveraging a modern ATS like Greenhouse.
  • Days from on-site to offer: Do you have a process in place to quickly gather interviewer feedback and come to a decision? Do we have a structured and consistent interview process? Again, a modern ATS like Greenhouse will facilitate this.
  • Days from offer to hire: Are candidates shopping around your offer? Are your offers competitive with market realities and candidate expectations? Are you rallying around the candidate after an offer has been made? One of the best interview experiences we observed was one where each member of the interview panel followed up with the candidate with a unique, personalized note to demonstrate their excitement.


With Time to Hire, your funnel metrics, your open requisitions, and your quarterly hiring goals, you can build out a simple forecast as to how you’re tracking towards your quarterly plan. It’s worthwhile to segment this by department or requisition so you can rally your company around hard-to-fill roles. Recruiting is a team sport, and a detailed forecast can illuminate to the team where their peers need the most help hiring. Consider using Tableau or a simple BI tool to build out a forecasting dashboard — Greenhouse lets you pipe data directly from a Redshift instance into Tableau or another system.

Candidate Resurfacing

This isn’t really a metric, but it’s worth mentioning. By the time startups are scaling, they’ve often seen thousands, if not hundreds of thousands, of applicants flow into their applicant tracking system. Most of these candidates lay passive in the system because they didn’t fit the role, or they went through some portion of the company’s interview process but didn’t proceed either because they weren’t a good fit, or they chose to go in a different direction. Startups can see value in resurfacing passive candidates in that are in their ATS — in essence, these are often candidates lying in plain site that are already familiar with your brand.

For example, a college student may have applied for an internship two years ago but not been a good fit, then she graduated and worked at another great company for two years, and now would be a good time to resurface her and reach back out because you’re hiring more full-time engineers with excellent backgrounds. Likewise, companies experience acquisitions, IPOs, shutdowns, lay-offs, etc. that affect candidate availability and can often be unexpected and rapid. Resurfacing candidates can be tricky but the easiest way is to tag candidates appropriately and set follow-up dates or a follow-up schedule that recruiters can manage to ensure they’re keeping passive candidates warm. Consider services like Atipica as well.

By Isaac Madan, investor at Venrock (email), and Shaurya Saluja of LEO Express. If you’re interested in discussing hiring & recruiting analytics further, please reach out.

Subscribe to our email newsletter here. Requests for Startups is a newsletter of entrepreneurial ideas & perspectives by investors, operators, and influencers. Email us with any questions or feedback.

Leave a Reply

Your email address will not be published. Required fields are marked *