SEMRUSH.COM
Organic research redesign
Significant product improvements
SEMrush started out in 2008 as a small group of SEO and IT specialists united by one mission — to make online competition fair and transparent, with equal opportunities for everyone. After ten years of trailblazing experiments and constant progress, SEMrush has grown into the world's leading competitive research service for online marketing.
>3
million users

>800
million keywords

>100
geo databases

>130
million domains

Team
Artem Gendler / Product Designer
Pavel Matveev / Product Owner
Objective
Organic research redesign
Significant product improvements
Timeline
Jul 2017 – Jan 2018
5 months
Context
I've worked with Domain Analytics these services help users to get instant insights into the strengths and weaknesses of your competitor or prospective customer. It provides information about traffic sources, top content, search performance, and many other.

Basically SEMrush analyzes the Google search result page and can provide reports about your site or a keyword. If you want to go into details or try this service go to SEMrush.

From a product design perspective, these services had a legacy design that lasts from 2008. And design team had a new UI-Kit so the service should be updated using this design approach.
My role
As a designer in a truly agile environment, I had no direct supervisor. So I had to align with my teams to understand their objectives so I could define my own objectives. Also as a designer, I should support design-related activities like updating UI.So I define for myself to workstreams:
UI update
Product design (excluding UI updating)
Discovery
I started from the current situation discovery. I conducted series of interviews with stakeholders to get known about business goals, team composition, known users pains.

Basically, products I worked with have some kind of shared monolith backend, that was overloaded and it was decided to create a super team with top company's experts to create a new version of this service. It was very challenging due to the 43 Petabytes of data used for reporting and the amount of Daily Average Users >100 000.

The next thing more related to UX is that users were often confused with system behavior in edge cases, they just didn't understand what's happening and what to do.

From the business point of view, these services were on top of the conversion funnel. So it was impossible to overestimate the importance of paid user acquisition improvements.

Based on the information provided, I was able to create a rough plan for the next steps:
  • Research (domain research, qualitative and quantitative user research)
  • Identify all edge cases
  • Identify all selling points
  • Gather detailed technical limitations (won't be disclosed)
  • Concept creation
Research
I started with domain research to gain an understanding of users before talking to them. So I talked to our marketing team and perform competitors analysis.

Afterward, we(me and the product owner) analyzed our usage data, session recordings, and customer feedback. So we had an understanding of friction points and were ready to conduct meaningful user interviews.

Product Owner reaches out to our loyal customers and we had a couple of conversations. And we understand that our respondents cannot help us much because they adapted so much to such behavior of the system that they could not even understand what we were talking about without a detailed explanation.

Together with the product owner, we decided to proceed with the data we already have.
Edge cases
From the user's perspective, all edge cases looked quite similar – there were very few items in a report or the worst-case report even couldn't be created. Most of the time that behavior was caused by the user's actions both conscious and unconscious.

As an example users search for https://www.welcomebc.ca and forget to switch the local database to Canadian. In the report, the user would receive just a couple of rows, so our newcomer user thinks that SEMrush is not so good as a research tool and left our service. As a result, we have lost our customers.

But if we highlight for a user that search was performed in the wrong database and provide a short path to the correct request this would dramatically improve user experience and reduce churn rate.
The final list contains 30+ edge cases.
Selling points
Business rules at those times looked like this we were limiting user possibilities according to paying plan:
Free (non-registered) users:
  • 10 reports allowed
  • 10 rows in each report
  • No advanced options like sorting, filtering, et
Free(registered) users, same as a free plus:
  • 10 additional reports a day

Other audiences were, as I recall, limited only by the number of reports.

When users reached their report limit pop-up contained credit card form blocks content. Users got confused by this form because they didn't understand what's going on. Also, most the users don't understand why there were only 10 rows in the report.
Concept
Creating concept I have 2 goals in my mind:
Remove friction in edge cases
Increase conversion


My solution was to provide useful information about what's going on and what to do.
Creating limitation concepts I was thinking about providing not only understanding of what's going on and what to do but benefits of paid plans.

As an example, I showed that there are a lot of results in this report, and highlight benefits. This approach increased conversion rate.
In the second example, we combined full advanced search experience with a friendly upfront reminder that this feature is available only on paid plans. This approach helps users to understand better the abilities of this feature. Also, an upfront reminder reduces the possibility of the user being frustrated by feature limitation.
Results
+ 10% conversion
Increased conversion to paying customers by 10% conversion.
- 30 % churn rate
Simple terms and explanations helped new users to rapidly fast order transportation services they want.
This approach showed very good results and these services remain unchanged from 2017 till 2020. Most important results:
  • New reusable approach for edge cases and selling points
  • Reduced churn rate
  • Reduced amount of tickets to support
  • Increased conversion rate.
Check Domain Analytics
Redesigned organic research
Made on
Tilda