From talent acquisition and employer branding metrics through to attrition rates, LinkedIn Talent Insights is offering a lot. But how solid is its data?

If you wanted an accurate picture of the workforce how would you go about attaining it? LinkedIn is betting that whatever your solution would be, it can do better. It’s created a product called Talent Insights (TI) with which it hopes to win over organisations who want data that helps improve its HR analytics.

The obvious advantage LinkedIn has over other solutions is that they don’t need to go out into the world to collect this data. Its users from across the globe are consciously and unconsciously giving the platform constant updates on their interaction with employment. From skill updates to which company pages they’re going to.

So the social network doesn’t just know when someone changes a job, it’s able to offer data on the macro and micro level. It can tell you how many software developers there are in Australia, and which company in Melbourne has been hiring the most. Would you like to know where your ex-employees go after the exit interview? Or the attrition rates of your three top competitors? TI promises an answer.

What’s unclear is the usefulness of TI’s answer. Its strength – that professionals on LinkedIn offer it data willingly – could also be its biggest weakness. And beyond that, is it a good thing that LinkedIn is opening up its data like this?

HRM went to the Sydney LinkedIn offices earlier this week for a product demo, in order to understand what the company was trying to accomplish.

HR analytics

What’s initially striking about TI is that if you were a company that had no in-company people analytics, simply adopting the tool would give you a start. It gives you a sense of how many people your company employed, along with each member’s tenure, seniority, skills, and so on (that is, so long as all your staff are on LinkedIn with up-to-date profiles, but more on this huge caveat later). You can also benchmark your analytics against industry standards – is your turnover in sales comparable with your peers?

Fascinatingly, it could provide you with a more accurate picture of why staff are leaving. If people in exit interviews are saying they’re departing over flexible work arrangements, you can verify how many ex-staff actually went to organisations that provide it.

One thing that might alarm some about TI, is you get the sense it’s too transparent. Yes, you can find out about competitors, but they can also find out about you. With a few mouse clicks, a rival can compare its workforce composition to yours, including where you’re getting staff from and what skills they have, and use that data to make strategic decisions.

When asked about this, LinkedIn representatives said this data was already available to those who really wanted it, and that their clients overwhelmingly feel the information benefits outweigh the negatives. It stressed that it puts member concerns (around privacy and other matters) first. The product has been out a while with no sense of a real backlash, but it will be interesting to see if one develops. Bigger companies might not mind, but will smaller companies – especially those with an owner-founder?

An example

In the LinkedIn demo, they gave an example of how they think the tool is best used.

Imagine you were an engineering firm pitching for business in Queensland and you wanted to scope out whether you should open an office in Brisbane. Your first question might be how many civil engineers are currently living in the city? By creating a talent pool report on civil engineers (the tool uses machine learning to cluster job-title variations) you could find out how many were in the capital, and then narrow your search down to the particular skills you might need.

TI can then tell you the relative demand of your ideal candidates  in the area, whether the total number is on an upward trend, which companies are the biggest hirers, and so on.

One of the more interesting features is TI’s ability to align this information with employment branding. LinkedIn does an annual membership survey that asks what people are looking for in a job, whether that’s flexible work practices or very competitive compensation. TI allows you to segment your talent pool using this data. You can find out how attractive you might be to Brisbane civil engineers by comparing what your company offers, with what they want.

It also allows you to search more broadly. For instance, if what you’re really looking for is really project managers that are capable of working with civil engineers.

Finally, you can easily do a similar search for other Australian or global areas, and find out where you could possibly be hiring from should a Brisbane office be essential and civil engineers are few and far between.

The caveat

As mentioned above, the big question mark hanging over  TI is that its data quality is very much dependent on the engagement of LinkedIn users. It’s not solely dependent, as it draws on information from off-platform in order to reduce inaccuracies (it scrapes recruitment platforms for duplicate job ads, for instance).

But it’s still mostly dependent. If a certain industry or profession is less engaged with LinkedIn, if the people in it are less likely to use the platform to make updates to their job status, or if they’re less likely to check out a company of interest, then the information TI offers about that industry isn’t as useful.

This might be especially true of job skills. If an individual sees LinkedIn as essential to their career prospects, they will make sure it lists every skill and experience they have. If they see it as tangential, they might not keep it up to date.

For instance, I used LinkedIn diligently after leaving university, when I worked exclusively in video. So my skills in editing programs and film production are highlighted. The only way the platform knows I have skills in news journalism (which I’ve spent far more of my career doing) is because of the endorsement of someone I’m almost certain I’ve never met (my apologies to Giovanmatteo from Universita’ Europea di Roma if I’ve just forgotten you).

The flip side to this is that LinkedIn has accrued a lot of data over a fairly long time period – which can provide hidden advantages. For instance, I still have video production skills, even though I don’t telegraph that fact, TI could reveal it. It could show companies whether their current employees have talents they didn’t know about, and so assist in hiring/promoting decisions.

Engagement levels

So what are LinkedIn’s engagement levels? According to Alexa, in terms of website popularity Seek (ranked 31st) is in the same ballpark as LinkedIn (20th) in the job hunting space. LinkedIn also has over 10 million Australian users (which is a lot, considering Australia’s workforce is about 12.6 million).

Of that number, according to Vivid Social, the platform has approximately 4.5 million monthly active users. Presumably all of those users have up-to-date profiles, so that leaves about 5.5 million profiles which would have varying degrees of accurate data. It seems promising, but the reality is that only LinkedIn has a full understanding of how engaged people are on its platform, and they’re not about to reveal unflattering statistics.

Would TI be valuable to your organisations? How solid is its data? If anyone has experience using it, good or bad, please comment below.