Algolia is one of the most successful developer-focused companies in the world.
It is basically a “great search box as a service".
Ok this is my explanation, sorry to all the product marketers at Algolia :).
I love their docs, incredible developer experience, minimalistic feel, and a super-powerful search engine product.
One snowy December evening I decided to understand how all the great people at Algolia approach developer marketing.
That evening ended up taking me a few weeks but I’ve learned a ton from it.
In this article, I will share all the research I’ve done along with the learnings (or vice versa).
Since this is a really long article here are the learnings so that you can jump into what is interesting.
Some of those learnings may not be as fascinating to you as they were to me but that is ok.
I just hope there are some developer marketing nuggets you will learn from this.
Ok here are the learnings:
- Learning 1: Search Hacker News - Algolia
- Learning 2: Powered by Algolia PLG button
- Learning 3: Free search box for open source projects
- Learning 4: Use developer-focused display ads providers
- Learning 5: Try advertising on Twitter an Reddit
- Learning 6: Run google search ads for typos
- Learning 7: Run ads to resources and examples
- Learning 8: "Something by X" product-led works like a charm.
- Learning 9: Another Product-led tactic. Free documentation search
- Learning 10: Another free tool, this time for GDPR
- Learning 11: When people write and company retweets you get more engagement
- Learning 12: Stream live coding sessions. Two devs are better than one.
- Learning 13: Dev-focused explainer video is about inspiring people to build
- Learning 14: Have your example pages easily searchable
- Learning 15: Have many navigational CTAs when communicating in-high level to devs
- Learning 16: Chatbot on home but not developer pages
- Learning 17: Show how it works very early on the home page
- Learning 18: Let people contribute and request examples easily
- Learning 19: You can gain trust without logos
- Learning 20: Create a no-sign up sandbox to “try before you buy”
- Learning 21: Add non-invasive “Need help?” button/chatbot
- Learning 22: A lot of subtle in-text CTAs and anchors that go from blog to docs
- Learning 23: Account-based marketing (ABM) is getting big for companies doing enterprise sales
- Learning 24: At Algolia scale and stage PR becomes important
I also found some things I didn’t quite understand which I mark as questions.
If you have ideas as to what it is about please let me know!
Ok, here they are:
- Question 1: Why not pay for other competitor keywords
- Question 2: Why use one landing page instead of customized landing pages per keyword (paid search)?
- Question 3: What Algolia started doing with their docs in May 2021 to grow it by 2x by the end of the year?
- Question 4: Why have two blogs and not just more categories?
- Question 5: Why did Algolia choose to do Public Discourse vs Private slacks/discord
- Question 6: Why have two similar pages for examples, not just one?
- Question 7: Why not have a CTA to build live in the sandbox?
Ok, let’s dive into this!
I ran a search on Dec/10/2021 on similarweb for algolia.com and got this:
As for traffic sources, it is almost an even split between direct 39%, referrals 28%, and search 29% with other channels way behind.
I often find developer-focused companies have way higher direct traffic and lower referrals.
The reasoning goes that as you grow way more of your users come from word of mouth and strong brand awareness.
Perhaps in the case of Algolia more people just go to search for "algolia" in google rather than type it in directly.
But why the referrals are so high for Algolia? Let's take a look.
So most of the referrals come from Hackernews, then Github, and a few lesser-known websites.
What is going on here?
Learning 1: Search Hacker News - Algolia
So the vast majority of referrals come from hacker news. Why?
I didn't understand it until I went into Ahrefs for SEO analysis and found this.
This is brilliant.
Hackernews is notorious for poor search experience and folks from Algolia created their own website where you can search Hackernews... with Algolia search engine.
Of course, when you click on "Search by Algolia" you get directed to the website and can learn how to set up a similar search, which you have just used yourself.
I love this tactic as it:
- solves a real problem for the audience Algolia is after (many software devs read hacker news)
- it shows rather than tells how Algolia search works. And it works amazingly
- it feels almost like an extension of HackerNews with the same colors and design.
And looking at the results it works brilliantly.
Learning 2: Powered by Algolia PLG button
A nice little product-led growth tactic is used for fontawesome.com where you see a little button "Powered by Algolia" next to the search box.
Many developers or people directly working with developers look for fonts so not only do people see Algolia when they use it but also those people are the correct audience.
A similar story is true for https://steamdb.info/, I suppose with a difference being that you catch devs in their free time as they are looking for games to play.
The reason why GitHub is coming up so high (I think) is that Algolia search engine integrates with many languages, frameworks, and libraries giving people a better developer experience when connecting their codebase to Algolia.
Learning 3: Free search box for open source projects
I think that another reason for Github being high on the list is that Algolia gives away their search engine for free to open-source and non-profit organizations in exchange for displaying "Search by Algolia"
I like this more than "just" sponsoring something or giving something for free to the community as it makes it easy for people to give back.
Just great search product
Socials don't seem to be a strong source of traffic for Algolia.
What seems interesting is that Facebook was so high on the list, followed by Type Pad (which I have never heard about) and Twitter hasn't even made the top 5.
We'll dive deeper into those a bit later.
Learning 4: Use developer-focused display ads providers
Looking at the pages that referred the most traffic to Algolia we see:
- codepen.io: social development environment for front end devs
- getbootstrap.com: front-end open-source toolkit
- materialpalette.com: a site that helps you choose a color palette for your site
Do you know what those sites have in common?
They are all frequently visited by the front-end devs which is great.
But also they are all powered by Carbon Ads, a display ads network that focuses on software devs and designers.
Have to say I didn't know about Carbon Ads but will definitely give it a go.
As I looked into those Ad Networks it seems that BuySellAds can also work well as the website says:
And apart from Algolia it is apparently used by LaunchDarkly and CircleCi (see case study).
Learning 5: They advertise on Twitter an Reddit
I also checked https://www.wappalyzer.com/ to see what tool stack Algolia is using and found this:
Just to get a sense of the style of display ad creative you can use Facebook Ads Library
First of all, nice domain authority (DR) probably makes ranking for whatever you write about easier.
A big part of the strong domain are referring domains (backlinks) and those seem every healthy at 26k and they grow steadily over time which means more strategy than round announcements and PR stunts 🙂
Looking at Anchor links that push to the website
I can confirm that this "search by algolia" button is working just fine.
Traffic seems to be growing slowly past few months but was sitting around 50k for most of the year.
Perhaps the fact that they started ranking for 30% more keywords in the last few months will be reflected in even more traffic coming their way.
Looking at the paid overview it seems that Algolia really wants to get people who look for "search engine" as this is a big volume and probably a very expensive keyword (and they got 31% of the traffic there). Only after a bit, I noticed that this high-volume keyword is actually a typo "serch".
Interesting thing is that the second most paid-for keyword was "flexport" which is a company and the link goes to the case study.
I don't quite understand just yet why they did that but it seems interesting.
The third one is a classic comparison page paid ad for keyword "elasticsearch", their main competitor.
What is interesting is that this "vs" page is very basic with only minimal information.
I would expect a deep comparison between tools, Gitlab style (see Gitlab vs Github). Perhaps it's an opportunity or they are already working on it.
Anyhow it seemed interesting.
According to Ahrefs, 5 keywords account for 99,56% of paid search traffic are:
- "serch engine": 30820
- "flexport": 2291
- "elasticsearch": 2171
- "algolia": 190
- "search sites": 57
Learning 6: Run google search ads for typos
And the first keyword is actually a typo.
Started typing some of the other things like "serch bar", "serch sites" and Algolia has live ads for those.
Elastic is and some other competitors are also paying for it.
Question 1: Why not pay for other competitor keywords
The competitor keyword "elasticsearch" costs quite a bit but it is probably still worth paying for it.
I was wondering why not go with other competitor brand names
Looked into Ahrefs and it seems that there could be some juice to be squeezed from it.
It is interesting to check which landing pages do they link to from ads and how those look:
- https://www.algolia.com/ : 450
- https://grader.algolia.com/ : 128
- https://www.algolia.com/site-search : 98
- https://www.algolia.com/search-engine : 34
- https://www.algolia.com/ecommercereport : 15
- https://www.algolia.com/search-api : 12
- https://grader.algolia.com/site-search : 11
- https://www.algolia.com/product/recommendations : 9
- https://www.algolia.com : 8
- https://resources.algolia.com/replace-endeca : 7
So the vast majority simply link to the home page which is against the common practice of setting up dedicated landing pages.
There are some dedicated landing pages, however (or so I thought).
Moreover when I went to see those landing pages:
Question 2: Why use one landing page instead of customized landing pages per keyword?
It is a common practice to make a connection between the keyword and the page with a customized value prop, case studies etc.
It is interesting to see that they had them in the past but dropped them.
So why not?
I figured I'd check what do the current (as of Dec/11/2021) ads for those keywords linked to and I found out that they all link to one page:
It's a nice landing page (we'll go deeper into the website in a second) focused on developers but it is very surprising to me that there only is one.
To confirm that I looked at Ahrefs Paid pages:
Yep, most traffic goes to this one landing page.
Maybe they decided to go back to the drawing board, establish a good baseline and build from that?
Or for their dev audience, it didn't really matter which use case as devs just want to build and you should focus on getting them to start building asap.
Either way, very interesting.
Nothing super interesting here IMHO apart from the "Ebooks" ad.
Learning 7: Run ads to resources and examples
Turns out there is an ad running that points to all the ebooks created by Algolia on their Resource page (more on this page later).
Apart from ebooks, there are also, videos, case studies webinars and so much more in there.
So it is likely people will get deeper into their buyers' journey and gain awareness about Algolia.
I am wondering if this is only for remarketing or cold as well.
As I typed "ebooks" into my browser I didn't get this as most likely the audience is somehow more targeted than me ☹️
Also, it seems that they are mostly targetting the US when it comes to paid traffic.
This is interesting.
There are 4 pages with 1-10k including 2 blog posts, docs page, and the pricing.
It feels to me that Algolia has a super strong brand in search (strong home traffic) and a lot of people are on the most aware stage of the customer journey (pricing).
Does it mean that ranking blog/docs is not that important to Algolia?
I don't think so.
Looking at the blog traffic over time it seems to be growing steadily.
What I like even more is that the keywords their blog is ranking for are veeeery relevant to the product which likely means that behind all this traffic there are conversions to signups.
This is a sign of a good, targeted content strategy and not just pushing for vanity metrics.
As for docs, the situation is more interesting.
Looking at the past 2 years for the first 18 months it was hanging at the same level and started to grow rapidly in the past 6 months.
Question 3: What Algolia started doing with their docs in May 2021 to grow it by 2x?
I tried checking the wayback machine for some obvious signals but couldn't see anything.
I think there is some gold in here but I don't quite get it yet.
Learning 8: "Something by X" product-led works like a charm.
Looking at Anchors again we see the "Search by Algolia" very high on the list with over 117k pages referring.
Figured I'd check what those pages are.
Wow, we have Tesco and Netlify at the top.
And sure enough, when I clicked on the search icon on the Netlify website I saw this.
Good job product-led team at Algolia!
This is a known trick but implementing this program at scale must have been a lot of work.
First I looked at the Top Keywords tab in Ahrefs.
There are a few interesting things:
- "crawler": they managed to rank with a product page for such a big keyword, nice
- "search documentation": another landing page, this time for a documentation search product from Algolia. This brings me to.
Learning 9: Another Product-led tactic. Free documentation search
I didn't know about it but Algolia offers a free search for public documentation pages on https://docsearch.algolia.com/
Another great move from the product-led team I suppose.
Obviously, software devs read docs a lot and great searching experience of the docs will not go without notice.
And when you search the docs of your favorite framework you will see the familiar "Search by Algolia" button.
Here are Gatsby docs for example:
Looked a bit deeper into those keywords as I wanted to see what the blog is about. Is it very focused on keywords that are relevant or is it geared more toward the general developer audience?
To do that I took all the keywords that the blog ranks for and gets 10+ visits a month according to Ahrefs.
Then I hand-labeled them for relevance to the product and stage of awareness.
So most of the articles are relevant with the exception of:
- One super popular article about git rebase
- Funding round announcement as apparently people are looking for "Algolia IPO" and "Algolia valuation"
Also, there are almost no product-aware keywords targeted on the blog.
That is very reasonable as you'd rather direct people to the docs with those queries.
So the docs rank mostly for product aware keywords but there is a healthy amount of solution-aware and problem-aware and even a little bit of unaware.
The more traffic for problem and solution aware you could get the more your docs can drive your signups and not only engagement.
Now the Top pages show a few more insights.
One of the top-performing pages is the classic SaaS marketing page: comparison between Elasticsearch and Algolia. I will dive into this page later.
Home page and Hacker News search in the top followed by the docs and the blog.
What is interesting, is that there are two blogs "/blog" and "/developers-tech-blog".
Question 4: Why have two blogs and not just another category?
Learning 10: Another free tool, this time for GDPR
Yet one more implementation of the same idea s Hacker News search.
This time for GDPR.
Nice follower count, simple description, clear banner. Everything is nice and slick as you would expect from a company valued at $2.25 Billion 😂
I like how they link to the page that was also used for the PPC keyword traffic that is a simple landing page focused on developers and how to get them building asap.
Let's dive into the feed.
The first thing that struck me was how little engagement people have with Algolia's feed. With this amount of followers, incredible product, deep relations with the open-source community I was expecting to see fireworks.
Yet most of the tweets have little to no engagement (I am talking 1 like 2 re-tweets).
The feed felt almost as if it wasn't meant for the devs but rather business folks.
And it didn't get much attention at all. Overall it just felt very bland, corporate and, I hate to say this, boring.
Now what are the tweets about:
- Promotion of the new solution pages, blog posts, industry reports, case studies, new podcast episode
- Some retweets of the customer, user, and industry tweets mentioning Algolia
- Add-like tweets about building search with and without video - I suppose it was used in ads actually.
- in 3 steps (with an awesome video) (I suppose those were ads actually
- Hiring shout outs
- Retweets of CEO
- Retweets of DevRels
Learning 11: When people write and company retweets you get more engagement
Overall the only tweets that actually got some engagement on the Algolia feed were written by real people in a conversational language and got retweeted by Algolia.
Very similar content to what was posted on Twitter. Directed at business and again mostly very bland.
It got more engagement than on Twitter but I didn't see anything really pick up.
The best things I saw were case studies that showed real results and some personal achievements of the CEO.
I am not a fan of the LinkedIn ads they run either.
For example, Who does this ad speak to?
Is it devs with "6 lines of code"? Or is it business with "30% increase in revenue"?
It doesn't feel like it speaks to either.
It is a nice design and all but I wonder who should actually care about this and click?
Same story as Twitter/Linkedin here. Little engagement, basic corporate feed, not much to learn imho.
Two things popped for me right away:
- Not as big following as I expected with 3.5k people
- A featured video that is not developer-focused but rather created for business/general public
Ok, let's look into the latest videos to see what Algolia folks have been creating recently.
Learning 12: Stream live coding sessions. Two devs are better than one.
I think before it was the Live Coding Sessions show that was moved to the new format.
I really like this and I think this format of 2 people talking and doing stuff is very engaging.
Learning 13: Dev-focused explainer video is about inspiring people to build
I liked this explainer video which was created for the dev audience.
- Minimalistic design with a little code and a little UI.
- Format of Step 1 Step 2 Step 3 Success that is focused on inspiring people to build
- Doesn't list all the features but goes through one path to success (and makes you feel like a person with superpowers)
There is an open community at https://discourse.algolia.com/
Question 5: Why did Algolia choose to do Public Discourse vs Dark slacks/discord
This is interesting that they chose Discourse while many dev communities form on closed slacks or Discord.
Dark social seems to be getting bigger and bigger and I wonder if the Algolia community stays public.
I like how “Log in”, the search box and the main CTA “Start free” are on the right side and the pages are on the left.
It makes the CTAs stronger and more visible.
Most pages in the navbar are business-focused but there is a special Developers tab where the majority of interesting (for developer marketing) stuff happens.
- Documentation and Hello world (Quick start guide)
- Examples (Code exchange, UI components, Inspiration library)
- Integrations (Client libraries)
- Community forum (Discourse community)
- Events (Sessions & live events)
- API status and Support
- Free, open-source option (For Open Source)
We’ll go deeper into that later but I think it is important to mention how many of these are examples.
There are three different pages where you can search through examples.
Learning 14: Have your example pages easily searchable
Have pages with examples where devs can find their use case or technology quickly.
Inspire them (get them to the aha moment quickly) so that they can start building in that inspiration flow.
Right off the bat, it is clear that this page was built for two reader groups: Developers who need to implement search and Business people who would like to add search to their websites.
This makes it super tricky to create a page actually.
Algolia deals with it nicely by making this home page a guide sign page to other places.
Above the fold feels like a classic SaaS product:
- Value proposition
- Visual that shows results
- Clear CTA
Very nice. Could be targeted both at engineers and business folks IMHO
It is followed not by a typical logo bar but by a list of use cases you may want to solve.
This page dispatches to use particular use case pages.
Below we have something for the enterprise.
The usual enterprise tick boxes: SLA, Security, Scalability.
Now for the connection between developers and business.
Not sure which obstacles, value points this crossover solves but it is a nice segway.
Below we finally have social proof, mostly business-focused.
What I really like are those links to deeper resources with particular use cases.
Below that we have a pure developer section focused on the most important part. How do I actually use this thing 🙂
Learning 15: Have many navigational CTAs when communicating in-high level to devs
Notice that CTAs are very subtle, not flashy.
They are almost like, “hey just choose your path” and include:
- Integrations with front end and backend frameworks
- Sign up (two options)
Typically conversion folks would say too many CTAs but I am not so sure.
If you focus on the developer journey navigational approach that gets people to what they want to see first before building is the way to go.
People want to see how much hassle it is going to be to build with it. This is one of the most important obstacles.
Finally, we have links to Algolia products, search and recommend. Nice and clean, navigation to deeper pages.
It is also important to mention that there is a Drift chatbot at the bottom right of the home and some other business-focused pages.
Nice and clean dispatch to different questions people may have.
Learning 16: Chatbot on home but not developer pages
The chatbot is not available on the developer pages which may suggest that it was considered spam by devs.
This is a bit extreme but I definitely see how this could be important for Algolia’s developer experience brand value.
It is the home page of the developer hub.
Algolia keeps it mostly navigational and educational, presenting the ease of use of the tool.
It jumps right to it, no value-focused stuff, it just tells you how to get things done.
Learning 17: Show how it works very early on the home page
With developers, you want to get to the point.
Assuming your audience understands what you do, jump straight to the point of how to get it done.
Once you understand how to do it in most frameworks it goes even more into integrations.
Then we have some links to examples of things built with Algolia that could help you get started quicker: Code exchange and Pre-build UI components.
Code exchange is a page with examples by use case which you can filter to find what you need.
It speeds up the time to hello world and makes it even easier to use Algolia in your projects.
Learning 18: Let people contribute and request examples easily
Right there below the examples, you have a box where you can request examples if something you need is missing or contribute one yourself.
I like the conversion trick used with the “Load more” button.
If all the examples would be loaded right away it would make it less likely for people to see the CTA.
Same idea with the front end components library https://www.algolia.com/developers/search-ui/
Actually, some of those things are exactly the same things you can find in the code exchange just presented a bit differently.
Question 6: Why have two similar pages for examples, not just one?
Then, they show some open-source projects build by Algolia community.
They show Github stars next to those projects to boost the trust I think.
We have two developer-focused product-led-growth marketing strategies presented below.
You get the Open-source search and Docs search bars from Algolia which you can put on your website in exchange for a “search by Algolia” button.
The product-led strategies that Algolia uses work really work as you know from other places.
A few latest blog posts.
Some events: live coding sessions, webinars, and stuff that you can see.
And some more links to important developer-focused pages.
Finished up with trust boosters for undecided.
Learning 19: You can gain trust without logos
There are no logos on this page. The trust is built with integrations and some reliability stats.
That was really interesting.
I wonder if that would be the case if the developer hub was actually the home page of Algolia.
Either way, I thought it was interesting.
Question 7: Why not have a CTA to build live in the sandbox?
Algolia has famously incredible documentation with amazing sandbox to play around with things to test the service.
It was really surprising to me that it wasn’t listed on this page.
Perhaps people who first visit this page go to the docs anyway but this sandbox experience is really incredibly good and I think it could boost some conversions a little bit if presented here.
Okey diving into Algolia docs in a short section would be really hard as those are just amazing docs.
But there are some great conversion tips in here.
The home page has a few components that I love.
Navbar with key pages that people go to and a huge search bar that you just cannot miss.
Big CTA to the interactive tutorial and if you don’t want that a video.
Learning 20: Create a no-sign up sandbox to “try before you buy”
Devs want to test things out with no strings attached.
And live tutorials, especially when they don’t have to set up anything, sign up for anything are a great way of giving them this experience.
Algolia goes a step further from a typical hello world example by giving me an option to choose a use case, code, and see the result right away. Brilliant.
Back to the docs, below that the header is a quick explanation of “What is Algolia” followed by a subtle Signup CTA.
I like it because many people may land on the docs from wherever and making it easy to sign up can actually push some conversions here.
Learning 21: Add non-invasive “Need help?” button/chatbot
If I need help I can always click “Need help?” and get it. When you click it it goes to the support page.
I like that it is not invasive but it is clearly visible and when I raise my hand to ask for help I know where I can do that.
It is interesting that there are actually two blogs:
It is especially intriguing considering that the main blog does have the “Engineering” category.
It starts with a featured post per category and a search box.
Very clean and non-invasive.
It is very similar to the category page by the way.
Then we have a few articles.
After a few of them, we see newsletter signup between articles.
Then after a few more articles, there is another crosshead with CTA to resources including examples, case studies etc.
It ends with another Newsletter signup box and a nice UX trick where you can search/load more articles.
The general article view is very clean and simple:
- a small search on the right
- social buttons on the left
- blog navbar with categories (always visible)
- regular company navbar which disappears as you scroll down
So where does the conversion happen?
Depending on the blog category the conversion strategy is different but for the “Engineering” which I presume is the most developer-focused...
Learning 22: A lot of subtle in-text CTAs and anchors that go to docs
There are almost no typically pushy CTAs like sidebar, horizontal bar, etc.
But there are just a ton of contextual anchor links to documentation.
I think this is absolutely brilliant:
- it doesn’t distract readers and keeps them clean and non-invasive feel.
- it only gets you there if you are interested in this particular concept
- it builds docs as a destination with a ton of anchor backlinks (perhaps this is how they managed to grow the docs SEO so much)
- devs like docs, especially Algolia docs where they can solve the problems they have right now
The most aggressive CTA I found was this “ *Just a general note on indexing....”
Nice search box with obviously visible “Search by Algolia” to showcase the product.
At the end, you get a CTA to sign up (in plain text).
Some related articles.
Get this thank you page.
This one is super minimalistic and doesn’t even have the Algolia bar at the top, just the link to the home page.
As for articles themselves.
Again, very minimalistic and clean.
Don’t really see any CTAs apart from in-text anchors to... you guessed it Algolia docs.
I looked at all people who had Linkedin profiles + open positions from the Algolia website to figure out what is the team structure at Algolia Marketing.
So it seems that there are separate organizations for:
- (Product-led) growth
- Developer Marketing
- Developer Relations
- Product Marketing
- Content Marketing
- Digital Marketing (including SEO, paid, Social)
- Account-Based Marketing (new organization)
- Field and Partner Marketing
- Customer Marketing
- Event Marketing
- Demand Generation (including lead qualification)
- Marketing Operations (including Analytics)
It is interesting to see how much specialization there is in a company of this scale.
What also stands out from reading descriptions on people's profiles and seeing their roles is that at this point it seems almost as if there were two completely separate motions:
- one targeted at enterprises, especially in e-commerce and retail
- one targeted at developers
Learning 23: Account-based marketing (ABM) is getting big for companies doing enterprise sales.
Seems like Algolia is getting into ABM which I assume from reading some public Linkedin descriptions.
It is an enterprise motion but at the scale of Algolia, it certainly makes sense.
Other big developer-focused companies do it already.
I got interested and found this video explaining how Snowflake uses it.
Actually, I found out that there are several social media manager and search engine marketing positions available so I guess folks at Algolia just don't have enough hands on deck to do engaging social strategy as well as do more than baseline on SEM. Makes sense.
Learning 24: At the Algolia scale and stage PR becomes important
They are hiring several folks for PR roles and it seems that at some point driving all the efforts to push your company beyond devs and into the public eye is really important.
Especially if you are at that stage getting closer to the IPO I suppose.
Also assuring the consistent brand story across all public communication is more and more important as you get bigger and there are just so many people communicating to the world about your company.
Thank you so much for getting this far!
I hope this was useful and you learned something that you can use in your developer marketing program.
If you have any comments about this or would like me to do similar research on a different company (maybe even yours?) let me know!