This article was created with the permission of Irina Shamaeva (Linkedin profile link ), one of the most influential people in the sourcing community. Irina also runs the blog and Facebook group Boolean Strings, the Internet Sourcing Community . I encourage you to follow the blog and join the group.

Irina has noticed that some LinkedIn features are not working or working incorrectly. She described it on her blog, e.g. in these posts: Job search is haunted and Linkedin software crisis .

Unexpectedly, she received a reply from the person in the position of “Director of Product Management from LinkedIn” that she would be contacted by a LinkedIn employee who would look at the errors described by Irina and answer her questions.

Irina was contacted by Dan, who works as an Engineer at LinkedIn.

Irina regularly shares communication with Dan on her Facebook group, which takes place via e-mails. Once again, I encourage you to follow the group , and I will summarize based on Irina's conversations and my own tests, what is currently not working or working incorrectly on LinkedIn . Let's start!

Years of experience

Starting with a reminder - the years of experience filter in the extended version of LinkedIn is not working properly.

What exactly?

  • For those with less than a year of experience, the filter rounds up to one year . It can be particularly troublesome when the candidate has several months of experience, and we are looking for people who have min. one year of experience.
  • People who have not entered work start and/or end dates are not searchable using this filter. The user below will be omitted from the search results.
  • If we do not have previous experience entered , only ongoing work or several jobs, we are excluded from the search results . Let us also remember about candidates who have not marked the end of their previous employment.
  • People with more than 30 years of experience are also not searchable (which is contrary to the information on the slider, but this is a "bug" of this filter).

Skills

Another example is the “skills” section from the search level.

Take a look at the two examples below:

Within standard search engines, we have information that "space" and "AND" express the same thing. This means that in our search in both cases we should have the same number of results, i.e. people who included the word “Python” and “Java” on their profile. How did we get a discrepancy in search results?

If we enter several words in the skill section, the filter will invisibly put quotation marks at the beginning and end of those words . Therefore, Python Java becomes "Python Java" .

To avoid the automatic placement of quotation marks, enter one of the logical operators, e.g. "AND" or "OR", or add quotation marks in the appropriate places yourself. However, we recommend selecting the skills that are available in the list separately in this filter.

Headline, summary and job description

Here the problem is mainly related to LinkedIn's basic search engine , but let's also be careful with the extended version.

Getting to the heart of the issue, words written in the headline, summary or job description are often impossible to find.

For example, I'd like to search for Hung Lee (if you don't follow his newsletter, I recommend signing up here ) based on his headline, which is "More signal, less noise, with Recruiting Brainfood" . By typing this phrase into the search engine, we will not get any results.

What is Dan's answer to this problem?

“However, the other reasons we don't index Headlines and Summary are more subtle. Firstly, the VAST majority of profiles (I don't have a number but it is likely above 90 or 95%) use “ at ” that LinkedIn provides by default. Basically, most people just choose the default, so it winds up being redundant. And the people that do use personalized Headlines usually (again, not always, but often) use pretty flowery language like “World changer” or sometimes slightly humorous things like “CEO at taking-a-break” or something. As a result, indexing this would wind up re-indexing the same information we already have, as well as indexing things that aren't critical.”

To sum up, headlines that are not spelled "job title at company" , e.g. "Tech Recruiter at Bee Talents" are often impossible to find .

Filter search errors

Below is a list of filters with occasional errors:

  • Industry - despite completing the industry section on your profile, we cannot search for this person through this filter;
  • Academic degree - a similar case as in the case of “industry ;
  • Company size - we do not know how companies that have not provided their size are classified;
  • Seniority - more information below;
  • Degree of contact - as above

Another filter - regarding seniority (included in the extended version of LinkedIn). It turns out that there are also irregularities there. At what seniority level do you think a person whose last job was related to a managerial position, as below, should be classified?

According to the filter, this user is classified as "senior".

Another example: the candidate's current job title is "Technical Lead". A role that is definitely related to management.

LinkedIn considers such a person as a senior as in the example above.

Why is this happening? The seniority filter is based on the length of time we spent in a given position. If we have more experience in a senior position and we are currently a manager (as in the example above), we will most likely be classified as a senior anyway.

The last filter discussed - " degree of contact ", i.e. an option related to showing profiles from groups 1., 2. and 3+, where theoretically the 1st degree are people from our contact network, the 2nd - people connected to people from our network, and 3+ - people connected with people with min. 2nd degree contact. In practice, however, it is a bit mixed up and when you search for people from the 3rd level of contact, people from the 2nd level are also displayed .

As you can see above, despite the “ 3rd Everyone Else ” flag in the filters, two people from the 3rd level of contacts are displayed, but also one from the 2nd level.

Job search

When posting a job ad on LinkedIn, keep in mind that your job offer may be difficult to find.

We don't yet know why keywords are not interpreted consistently in Job Search.

Why are there likely to be no new filters or features?

Quote from Dan “It might sound weird, but each “field” we index and maintain costs us. It costs us in engineering, and also in speed. We spend a lot of time trying to make Recruiter (LinkedIn) FAST, and everything we index frequently makes it slower, by definition. And not just slower now, but slower over time (because we add more and more members) .”

The answer is mainly that adding new filters would slow down LinkedIn. Let's hope that will change someday.

Summary

At the moment, I gathered the most important information after Irina's conversations with Dan, an Engineer working at LinkedIn.

The number of errors is not small. It is possible that in the proverbial "in the meantime" we will learn new information from Dan, and some functions will be improved (or other bugs will appear). So I will try to update this post when it happens :)

Maybe you see other errors related to searching on this platform that did not appear in the post? If so, leave a comment or write to me on Linkedin ! :)