The client needed to reach developers in new markets in Asia and Latin America with a story about remote work, but first it had to understand what they read. We skipped the rankings and asked the developers themselves. Their answers turned into a campaign, and the most valuable findings were in the details, such as where local media barely reach developers and why one article got several times the views of another.
Results at a glance
Impressions of the banner
~300,000
A tech publication in Latin America, per the publication’s own statistics.
Views of the first article in Asia
3,000+
A tech portal. The article spent a day on the homepage.
Impressions come from the publication’s statistics, and views from the counter on the article page. Under the terms of our contract we don’t name the client or the publications, which is why the figures are rounded. There’s no search traffic to report here, because this was media research and PR with no SEO component.
Before
Two markets and no ready-made media list
| Client | a global IT services company, engineer hiring |
| Niche | remote work in tech, employer branding |
| Markets | new markets in Asia and Latin America |
| Services | media research, negotiations with publications, campaign management |
| Period | 7 months, 2022 |
| Status | completed |
| Team | an outreach and PR specialist and a project manager |
The client hires engineers to work remotely. People apply, go through interviews and tests, join the platform and then work on client projects from wherever they like. We’d been running SEO for a new section of this client’s site for more than six months when the hiring team came to us with a different task, about media.
The client needed to tell experienced developers in new markets about remote work. What was missing was an answer to the main question, what these people read.
The goal
Show up where developers actually read
The goal was to reach the right readers, and the constraints were these:
- experienced developers only, so platforms for juniors are out, however much traffic they have;
- neither we nor the client had a ready list of media for these markets;
- the client made the decisions, and it expected facts from us to make them.
What we did
We asked people first, then negotiated with the media
Seven months of work fall into three stages.
Month 1
Nobody had a ready list of media, so the developers themselves gave the answer, and within three weeks there was a vetted list
There was no ready list of media for these markets. We went to developers on LinkedIn and Facebook ourselves and asked them, vetted every site against the same criteria and moved on to negotiations within three weeks.
Asking the developers themselves
Big traffic shows what’s popular in general, while the client was after an experienced developer, and where that developer gets news and job openings is something they know better than any analytics tool. So we searched four ways at once:
- our own contacts;
- relevant Facebook groups;
- direct search, checking each site’s audience in SparkToro and Similarweb;
- over 40 LinkedIn connection requests to developers in both markets with one simple question, what they read.
In the analytics tools we looked at each site’s main audience country and at direct visits, because those mean regular readers who open the site themselves.
The same criteria for everyone
A well-known name doesn’t mean experienced developers read it. So we vetted the big, familiar platforms against the same criteria as our own finds, and sites aimed at juniors and career switchers dropped out, however much traffic they had. All the media went into one table with the same parameters for each and a column answering the client’s questions.
Real terms come from the correspondence
A rate card on a publication’s site is only a shop window. What you actually get for the money (whether the article reaches the homepage, whether it goes into the newsletter, how many people will see it) comes out in the emails. So before the whole list was even approved, we wrote to several publications in both markets and asked for reach data and wider business programmes alongside the price.
What the client did. Approved the first sites for outreach and asked us to move from searching to negotiating.
Outcome. Within three weeks we had put together a vetted list of media, and the client approved some of them for outreach.
Month 2
Negotiations showed what the media really sell
Media terms only came out in negotiations, so we asked about reach alongside price, talked to editors in their own language and put the terms into documents for the client’s decisions.
Asking about reach
A cheap placement nobody reads makes no sense at all, so we asked each site about reach alongside price. It turned out that big portals give guaranteed views only in pricier packages, so basic articles are worth testing one at a time.
Negotiating in the editors’ language
Big local publications may answer an email in a foreign language with a template or simply send it to spam. So we talked to the Latin American media in Spanish, and that’s how we reached the corporate accounts representatives of the big publications.
Documents for the client’s decisions
The client was choosing between sites with different currencies, taxes, formats and guarantees, and the terms were scattered across dozens of emails. So we put each site’s terms into a separate document and all the media into one table.
What the client did. Chose the sites for the campaign.
Outcome. Within a month the client had prices and reach figures for the sites and had built a campaign.
Months 3–7
The banner got about 300,000 impressions, and the homepage decided how many people read the articles
In short, each publication played by its own rules. We agreed the editors’ changes while keeping the core of the message, and each creative and each article led to the client’s site with its own tags. Article views depended on whether the article made the homepage.
The editors’ changes
A local editor knows their readers well, and the editors pointed out which wording would sound too promotional to them. So we agreed the editors’ changes with the client while keeping the core of the message.
Separate tags for each creative
Without tags, the clicks from a campaign merge into one stream, and you can’t tell what worked, the banner or the article. So both banner creatives and each article led to the client’s site with their own UTM tags.
Do they read local media at all?
A list of sites the audience doesn’t visit would give the client nothing. So in new markets we first went back to developers on LinkedIn and asked whether they read local media at all.
What the client did. Provided the texts and a banner in two creatives, and approved the edits.
Outcome. The banner on a tech publication in Latin America got about 300,000 impressions. The first article on a tech portal in Asia spent a day on the homepage and got over 3,000 views, several times as many as an article without the homepage. In some markets developers barely read local sites and talk in Facebook groups and Telegram channels.
What was hard
Every publication played by its own rules
The homepage decided the views
The article that spent a day on a tech portal’s homepage got several times the views of the one without it. Big portals give guaranteed views only in pricier packages, so basic articles are worth testing one at a time, with that built into your expectations.
Editors reworked the texts for their readers
Local editors suggested their own changes so that the text would sound natural to their readers. The client’s goal was our limit, so we agreed the changes with the client and kept the core of the message.
After
The client knows where its audience is, and where it isn’t
| Before | After |
|---|---|
| No ready-made media list | A list of media vetted against the same criteria, some of which the client approved for outreach |
| No idea how many people would see a placement | A banner with about 300,000 impressions and articles in vetted publications |
| No idea whether developers read local media | Developers told us themselves that in some markets they talk in Facebook groups and Telegram channels and barely read local sites |
In this project the target action was awareness among experienced developers in these markets and visits to the client’s site from a banner or an article, so each creative and each article had its own UTM tags.
Why it worked
Three decisions that made the media plan accurate
We asked the audience
More than 150 developers in several markets got a simple question from us, what they read. Their answers produced a list you won’t find in any ranking.
Same criteria for everyone
We vetted the big, familiar platforms the same way as our own finds. That’s why sites for juniors and career switchers dropped out, however much traffic they had.
Reach alongside price
What you get for the money comes out in the emails. So we asked each site about reach and put the terms into documents for the client’s decisions.
Takeaways
Four lessons for anyone hiring abroad
- Ask your audience before you buy media. Here over 40 LinkedIn requests produced a list of media for both markets that you won’t find in any ranking.
- Check whether your audience reads local media at all. In some markets, developers told us outright that they barely read local sites.
- On a news portal, views depend on placement. The article on the homepage got several times the views of the one without it. If placement isn’t guaranteed, build that into your expectations.
- Trust the local editors’ changes. A local publication knows its readers and can suggest how the same message will sound more natural to them.
Questions
What people ask about this case
How do you find media for employer branding in a market you don’t know?
We searched four ways at once (our own contacts, relevant communities, SparkToro and Similarweb data, asking developers directly on LinkedIn), and the survey produced the most accurate list.
Is the cheapest package on a big portal worth it?
As a test, yes, as long as expectations are right. On a big portal a basic article doesn’t go into the newsletters, so its views depend on whether the editors put it on the homepage. In this project the gap between two articles was several-fold.
Do you do PR without SEO?
Yes. This project was media research and negotiations with publishers, with no SEO component. At the same time we were running SEO for a new section of the same client’s site, which was a classic search project.
