Learning how to understand your own people and fight for them with Yandex
Choice is good. With choice, the possibilities are endless. — Arkady Volozh
What comes to your mind if somebody tells search engine? With high probability, it's Google. Right? Google has been dominating in so many major areas in the worldwide internet. It all started with Search but have had enormous growth with Video, Map, E-Mail, … and now even Google Home, wearable technologies and driverless cars. To be honest, majority of us all love Google products which are monopoly in our lives. In the markets where monopoly exists, users don't have either much or no choice at all. In other words, users have to follow and use only one major service.
But it's better to check markets where competition is high. Where each player in the market pays much more attention and efforts to user needs. In other words, users will benefit from the competition since they will have chance to choose one of the choices with respect to their needs. Not just because it's the only choice.
I remember the first time that I heard "Choice is good." It was a bold move from Yandex, a Russian Search Engine giant. Arkady Volozh, Founder and CEO of Yandex was telling: "Choice is good!" It was enough for me to be convinced why a competitor is here against big brother, Google.
Yandex was coming to Turkish market as it's first real international experience. A market where Google was dominating with almost 99% market share. Besides all the numbers, there were so many challenging roadblocks ahead:
- A new language and alphabet rather than Russian and cyrillic alphabet
- Strong belief and faith in the competitor
- A totally different user behaviors and needs
But as I said, it was a bold and exciting adventure we started. Everybody was aware that it's not going to be easy. Yandex needed three immediate things: understanding Turkish language and user preferences for Search quality, localizing Yandex products and creating the brand image for Turkish users.
It was a time when Yandex left it's comfort zone and started a war.
Search quality improvements
You may have a powerful system running the search engine. It can be so smart and providing excellent results and satisfaction for users where you know and already operate. But how about with a new language and alphabet?
One of the main strong points of Yandex is understanding Russian queries and crawling Russian documents. If you understand the query properly and have a huge index of documents, you will be able to match each user needs with your search results. This skill also brings success in Post-Soviet States, since Russian language is still affective there. But the real challenge was understanding Turkish with every single detail and letting machine learn it properly.
These improvements were unseen heroes of Yandex for Turkish market. In order to have a better search for Turkish users, millions of queries are collected, documents are analyzed and continuous tests had been held. A huge team was working on Yandex search quality and improvement. And even we as Turkish people were quite surprised by users real intentions. It was a great amazing opportunity for us to understand our own people.
Search quality doesn't mean only clicking the top item on the search result page:
- Do users come back right after they click?
- In which page do they find relevant result?
- Do users come back to Yandex to make a search again?
- How much time do users spend in the documents they clicked on your search result page?
- How many query variations do users try to find appropriate results?
- What are the synonyms for different regions of Turkey?
I am considering to write a separate post (or posts) about how search engines evaluate their performance and how they improve it. But after 2 years, we were happy to see the comparison results of Yandex and Google. Comparison tests were held by independent research companies, and they were sayin Yandex performing much better than Google with respect to user preferences! We were also monitoring increasing number of users who are leaving big brother and coming to us.
ps: Best part of analyzing user intentions was seeing consecutive queries. Like, which queries are usually searched after first search. One of my favorites:
- e-okul (e-school grade system)
- e-okul login
- how to hide grades in e-okul
- how to change password of parents for e-okul
- what should I tell my parents for my bad grades
Funny, isn't it?
Localizing Yandex products?
Many people would call Yandex as a search engine company. And that's right. However, Yandex provides many other beneficial solutions and services to users. Like Yandex Maps, News, Tickets, …
So while entering Turkish market, Yandex decided to localize all products of theirs which are already successful in Russia and Post-Soviet countries. But localizing is never enough to get the same success. If you are ever in such business, first ask yourself following questions:
- Is user experience in the new market similar to our existing markets?
- Are these services really needed in the new market?
- Are these services worthy to focus and invest in the new market?
Basically, translating services does not (should not) mean localization. That's why most of the Yandex services didn't really match with user needs in Turkish market, except Search and Maps.
Even though, Turkish market needed (still needs) some of the Yandex services a lot, probably the timing or the implementation of ours into the market was wrong. For example, Yandex.Market is an amazing service that may help users while finding appropriate goods with fitting prices.
On the other hand, not after a long time we figured out the way of success after having some failures. Keep reading, you will find them out!
A new brand: Yandex
So, as you are entering a new market you need several mandatory things. But following two of them must be there to exist:
- Product(s) or service(s) to help users
- Brand recognition
So therefore our marketing team was focusing on how to create Yandex brand in Turkey and sustain it. Yandex entered Turkish market so fast with aggressive approach. And we were attending all conferences, ad videos, TV commercials… Yandex was everywhere. And even the slogans that we were using in the commercials became Twitter trending topics!
Ps: Maybe one day one of my ex-colleagues in Yandex Marketing team will publish a story about their work on this topic too.
Here are some examples of our TV commercials. To be honest, some people in Turkey was thinking our commercials were not so good and but all of them were so catchy.
But, wait a minute!
But something was wrong. Even though we spend so much efforts, we couldn't get exact return on our investments. Users were not fully convinced to moving to Yandex. So we kept asking ourselves, what's wrong?
We came to the market with a powerful search engine which has competitive quality over the Google, maps, video, and many more other services. But still, most of the users were trying Yandex, but we were not able to retain them properly.
It was time to question our approach to the users actually. And we asked one crucial question:
Do we really know what Turkish users need and like?
Maybe we were missing important points and offering our products either in too generic way or with wrong offerings. Users were probably skipping our promotions just because they are not offering:
- to solve any of their pains
- to bring any kind of gains
- to complete the missing part in their online needs
Let's listen users for a second
After we asked most crucial question to ourselves, our main action was so clear: Let's talk with Turkish users.
If you are doing a B2C business, never and ever miss the chance of talking with users! You never know what's hidden in their daily life and you cannot find most of them by analyzing their usage data.
Before starting such process, I'd recommend to decide followings. But you may also change them during the process in case your assumptions don't really match the situation:
- Strategy: What you aim from the process
- Technique: How you do the process and do the interviews
- Segmentation: Who you are going to talk during the process
- How to get and use output: In which format you would like to get output
Our strategy was starting understand user needs from the top point: how they use internet in their daily life. We believed and knew that, if we start asking questions regarding Search we may miss many important points, users motivations, their daily internet usage limitations, etc…
And for the technique, we followed in-depth interview and worked on it. There's an upcoming post about what's in-depth interview and how to do it! Basically, in-depth interview is the most common qualitative research method. It's important to focus on techniques getting good results or not to ruin the interviews. Participant should feel comfortable and being part of a chat instead of just answering some questions. This technique brings the possibility of getting deep down thoughts during the conversation aka interview.
For sure, it's really important with who you are doing interviews. While planning the whole process, you should have an idea for the participant candidates. I'm expecting everyone who'd like to do such think to know: User profiles and their demographics with shares. While first starting interviews, you should start with generic segmentations without narrowing users too much. As you start going into deep for specific areas, you will have chance to identify your desired segments better. For example, you may define your segments step by step like this:
- General Overview:
- Men / Women, 18–40 age group
- Who uses internet at least 3–4 times a week
- Who makes search at least 2–3 times during days when using internet
- Focus on video search
- Men / Women, 18–40 age group
- Satisfies first segment group
- Who watches videos on either desktop or mobile devices at least 1–2 times a week
- Who makes searches to find the video
- Focus on tv series search
- Men / Women, 18–40 age group
- Satisfies second segment group
- Who watches videos of TV series on either desktop or mobile devices at least 1–2 times a week
- Who makes searches to find the video
- Who follows up TV series up-to-date
These segments are given randomly. Each step segment requirements depends on the output of your previous step. Segment example is given with respect to you found out users are really interested in searching and watching videos online, and then they have a huge interest to follow tv series up-to-date online rather than TV.
Besides all these technical explanations, let me go back to the story part. As we had these interviews, our world on products was enlightened! Users gave us so many Wow affects where we should have did that with our products. I'm admitting that Turkish users prepared their own roadmap for us.
This whole process took months for us, but we were in a great agile environment. We were applying all new learnings into new product development or improvement of already built features. Results had to be great since we started to see users:
- Untold desires
Then it was time to give them what they want.
Give users what they actually need
So we discovered all these areas we should take an action. Instead of fighting against with the competitor, we found the clean waters to sail our ship. As described in Blue Ocean Strategy, don't go to the bloody red oceans. Find your undiscovered blue ocean where you will benefit without competition. In other words, we discovered many topics to focus where users have need and competitor is not satisfying either.
And it was much easier to convince different product teams to work on new ideas since we had so many reference points. Since we were inviting each related product team members to the interviews, they also had chance to learn with us.
OK — You know what to build. You build them. Then what? You need a good channel to distribute your product. As a search engine, ultimate goal is going into the users' browsers. And never do that without forcing or doing it secretly. Users will hate you for sure. (As they did for Yandex during first years)
So we built a conversion channel for users. Aim was bringing users from competitor with respect to their needs by using ads to us. And users were asked to install our search to their browsers with their own will. And results were so good. Conversions were high and retaining users were also high. Why?
Because we were providing real needs of them.
Voilà! Here we are…
And with the new methodology to improve our products and distribution channels, Yandex's usage and active users were increasing enormously! We achieved 9–10% market share just for search. Meanwhile, Yandex.Maps and Yandex.Navi were already dominants in map products.
Yandex has been called as Turkey's Search Engine.
Was it a luck?
- Absolutely not.
What was the trick of ours?
- Never giving up. And testing more and more ideas.
What did Yandex bring to you?
- Chance to learn our own people deeply and their needs
- Achieving something most of the people didn't believe
Choice is Good for You. Ask for it.
Thanks for reading such long post which sheds light on very small part of the whole story!