Grammarly, a popular tool offering AI-powered communication assistance, was founded in 2009 with a mission to improve lives by improving communication. In time, the product has scaled from individual, subscription-based writing support to Grammarly Business, which helps organizations stay on-brand and accelerate business results through effective communication. In 2021, the company introduced Grammarly for Developers, which enables programmers to embed Gramamarly’s text editing functionality and writing suggestions into any web application.

“Gem gives me the data I need to iterate on sourcers’ volume and the quality of their reachouts. That’s the most impactful thing we look at right now from a performance-management perspective.”

Andon Cowie

Head of Global Recruiting

A partnership to integrate its technology into Samsung’s keyboard marked another milestone on the company’s path to help people communicate effectively wherever they write. In November 2021, the company raised $200M at a $13B valuation, making it the 10th most valuable U.S. startup. Grammarly now serves over 30 million people and 30,000 teams daily. “Talent acquisition has been buzzing and vibrant for as long as I’ve been at Grammarly,” says Andon Cowie, Head of Global Recruiting at Grammarly. “And that doesn’t look like it’ll slow down anytime soon.”

“We’d used Gem at previous orgs—mostly for automation”

Both Andon and Dana Schafer, Recruiting Operations Manager at Grammarly, used Gem in previous roles at other companies—primarily for its automated sequencing and the time-savings it afforded. “We brought Gem in at Dropbox to help engage candidates. The automated follow-ups saved our sourcing team a ton of time,” says Andon. “So when I came to Grammarly and looked at our tech stack, I knew Gem would be a critical addition. I implemented it within the first few months I was here.” Dana was previously in a full-cycle recruiter role at Pinterest, “and I used Gem for outreach sequences there. It was exciting to get everything out of spreadsheets, to make sure I was actually following up with prospective candidates—that there was process and rigor there. When I arrived at Grammarly, it was a great surprise to see that Andon had implemented Gem. I never wanted to do outreach or tracking manually again.”

A step up: from time-savings to self-service performance optimization

In their respective roles now, Andon and Dana still appreciate the power of Gem’s outreach automation and metrics, which allow sourcers to personalize outreach at scale, A/B test, send-on-behalf-of hiring managers or execs, and understand recipient behavior through opened/replied/interested rates as well as content stats. “A huge benefit Gem offers is in our top-of-funnel sourcing efforts,” Andon says. “It gives me the data I need to iterate on sourcers’ volume and the quality of their reachouts. That’s the most impactful thing we look at right now from a performance-management perspective.”

What’s great about the team-wide visibility Gem offers, Dana adds, is “we don’t necessarily have to use it at a manager level. Gem helps individual recruiters proactively understand their own performance. A sourcer or recruiter can go in and look at how their funnels compare to related funnels. All the data is there to answer the questions: How am I faring in my outreach? What’s working and what isn’t? How do I iterate and optimize to discover my own best practices, and ultimately to see better response rates?”

Coming prepared to meetings with leadership

But what’s been equally important are the high-level metrics that help Andon and Dana better communicate where recruiting stands with leadership. Gem’s dashboards feature, which organizes and displays the most important recruiting metrics for execs, has saved Andon hours in collecting data, made certain meetings irrelevant, and allowed him to preemptively catch issues that he can quickly bring to the attention of the right leaders.

“I used to go to an ops team and an executive team meeting on a bi-monthly cadence,” he says. “It would take the morning to pull all the data from various sources—our ATS, internal spreadsheet trackers that recruiters updated manually—into a single document. I was pulling basic data: hires per function, accept rate by department.” Now, Andon says, those meetings aren’t essential, “because the data is right there. Execs can self-serve it. Plus I have access to data I didn’t have access to when I was manually cobbling those reports together—average time to hire, for example, or offer decline reasons, or offer-accept rates by gender or by source.”

“There’s always going to be some issue; that’s a given. The question is: how quickly do you want to identify it?”

Having that data in real-time, Andon adds, helps him think well beyond the question of: Are we on pace for the quarter? In larger organizations in particular, “it can be easy to lose track of number of hires by function. With Gem, that data is immediately accessible. But we’re also doing one-better: we’re identifying issues before they’re even issues. For example, we recently caught a modest spike in decline rates for a certain function. I spotted it in real-time through Gem’s dashboards, got ahead of it, and brought it to our exec team to pivot so that trend didn’t continue. You observe imminent challenges or obstacles before they’re on top of you.”

Andon and Dana both stress that—as Andon says—“the common theme is there’s always going to be some issue. The question is: how quickly do you want to identify it, and how quickly do you want to understand its urgency? A single data point can quickly escalate to a larger problem that creates a ton of work for recruiters, recruiting leaders, and hiring managers—because it happened once. The bird’s-eye view in Gem puts it into perspective: maybe we look at the data and see this was an anomaly. That saves us a ton of time in tackling problems that we ultimately don’t need to tackle, in prioritizing the right things. We can revisit if it comes up again in Gem.”

That ability to prioritize is “probably the most valuable thing I get out of Gem in my role,” Dana adds. “It would be easy to look at high-level data and say, There’s a problem, we have to overhaul everything, code red! But with Gem, I can get granular and slice the data in a number of ways. If something looks off, I can drill down by department, by hiring manager, by candidate demographics, by location. This is what’s happening in North America; this is what’s happening in Ukraine. Suddenly I have a more easily-scoped-out problem that doesn’t feel as overwhelming or as wild. What’s working in Ukraine that isn’t working in North America? Rather than an entire process overhaul, it’s a diagnosis of the true pain point, and an application of best practices from one location to another.”

“In Gem, real-time customization is just a matter of adding and removing widgets”

That’s where Dana’s role comes in, she says: “to observe and then pass that information on to the leaders who know those domains best and can influence change management there.” And because what’s top-of-mind for Dana is in constant flux, the customization feature in Gem’s executive dashboards has been invaluable.

“What’s top-of-mind for me now is time-to-hire,” she explains. “So thanks to Gem’s customization options, I have a lot of real-time widgets that show offer-to-hire in Ukraine, offer-to-hire in North America. I can compare by function. But six months from now, it’ll be something else, and I’ll want a different set of views. In Gem, it’s just a matter of adding and removing widgets. I can also make audience-specific dashboards, because the data that’s critical and actionable for recruiting leadership isn’t the same data that’s critical and actionable for our business leaders or hiring partners. So I add and remove widgets. It’s really that simple.”

That customization feature is so important, Andon adds, “because the metrics we were looking at a year ago, two years ago, aren’t the same ones we’re prioritizing now because of our scale. Quality of hire was much more critical than time to hire a few years back. We didn’t care if it took a hundred days to hire someone, as long as it was the right person to come in and build early with us. We still want to find the right people, of course; but it has to be scalable. As Dana alluded to, one of our OKRs going into next year is time to hire. So we’ll customize the dashboard, and we’ll see that first-thing in the Gem view.”

End-to-end funnel visibility: using source-of-hire data to influence where the team spends their time

Another critical metric for the Grammarly team is source of hire, “which has influenced a lot of our big initiatives this year,” Andon says. Having source-of-hire data in the dashboard ensures that the team isn’t going on gut instinct or anecdote, neither of which is reliable intelligence. “For example, there was an impression recently that one of our locations was having trouble sourcing. It was a story the team had bought into—that the sourcing function wasn’t filling pipelines. We went into Gem and looked at real-time data. It turns out that sourcing is the best source of hires for that location.”

What was really happening, Andon says, “was a capacity resource problem that made it look like the team wasn’t filling pipelines. There was actually a ton of activity happening on the sourcing front, a huge knowledge set there.” Andon says that this intel has helped influence where the team spends their time at the top of the funnel, which is “an enormous challenge for most companies right now. Plus, getting to see that disparity between perception and truth is fun.”

“It’s fun to track, to be honest… but it also ensures data integrity.”

Andon stresses that Gem offers the best of both worlds: “the ability to answer questions in real time for our executive bench, and the delight of data-driven results, which is what should keep you motivated as a talent leader.” From an ops perspective, Dana adds, Gem ensures data integrity. “Data integrity can be tricky because you don’t always know the data is wrong until you’re pulling a report,” she explains. “At that point, it’s on you to check in with recruiters, go back in, and clean it manually.” But in Gem, “if something shows up unlabeled, it’s right there in front of me. I can identify it, go have a coaching moment with a recruiter, and ensure they’re inputting data correctly going forward. I don’t think we talk about the importance of data integrity often enough. But it’s essential if we want to really understand our process.”

Andon points to the offer decline reason widget in Gem’s dashboards as an example. “Prior to Gem, recruiters would just select ‘General’ or ‘Other,’ which didn’t help us understand the story behind a specific offer decline. At the end of every quarter we’d have to go back and do an audit, check in with individual recruiters to see if they could remember whether ‘Other’ meant ‘wrong timing’ or ‘cold feet’ or something else entirely. That ultimately didn’t provide us with good data, so we couldn’t make quarter-to-quarter shifts based on real intelligence.”

Now, Andon says, the moment he sees an “Other,” he can check in and ask what the unique reason was for the decline. “What that means is real-time fixes and pivots. If career growth is an issue, I can say, Hey managers, make sure you’re talking about growth opportunities with candidates from the very start of the conversation. When we have a clear idea of why candidates are opting out, we have a clear idea of what to lead with from first contact.” Offer decline reasons in Gem help Grammarly become a more competitive employer, because the team now knows exactly what to focus on to positively impact offer accept rates.

Finally, Dana notes that the talent acquisition team at Grammarly has always had “really solid partnerships” with their executive team. What Gem has added to those partnerships is “an enhanced level of trust. When something is being updated in real time,” she says, “leadership knows for certain that there’s nothing that’s not being shared. We’re all on the same page with the same visibility. And we all get to be proactive—as opposed to reactive—with the intelligence Gem shows us. It truly makes the ongoing growth we’re seeing at Grammarly a team sport.”

“It would be easy to look at high-level data and say, There’s a problem, we have to overhaul everything, code red! But with Gem, I can get granular and slice the data by department, hiring manager, candidate demographics, location. Suddenly I have a more easily-scoped-out problem that doesn’t feel as overwhelming or as wild.”

Dana Schafer

Recruiting Operations Manager