Templates
Removing the adoption barrier
Templates
Removing the adoption barrier
Templates
Removing the adoption barrier
Overview
Overview
Overview
As it's a very dense and complex feature, we face some struggles with adoption. Some users find it complicated and difficult to use.
In order to make this feature more friendly and remove barriers to entry for less tech-savvy users, templates were first introduced in 2018. This was proven to be a major adoption driver, resulting in 44% of automations created from templates.
The first iteration of templates worked well enough, but as Pipedrive grew in functionality and more integrations were introduced, we needed to revamp the templates in order to integrate them with our updated codebase, make them scalable and future proof.
My contribution
My contribution
Product strategy
User Research
User Interface
Product Management
The team
The team
1 Product manager
1 Front end engineer
1 Back end engineer
1 UX Writer
Year
Year
2023
2023
Context
Even though there were some technical requirements that helped this project to gain priority in our roadmap, the main problem was quite simple:
Users think automations are too complicated and don't know how to start automating their sales processes.
Context
Even though there were some technical requirements that helped this project to gain priority in our roadmap, the main problem was quite simple:
Users think automations are too complicated and don't know how to start automating their sales processes.
Context
Even though there were some technical requirements that helped this project to gain priority in our roadmap, the main problem was quite simple:
Users think automations are too complicated and don't know how to start automating their sales processes.
Discovery
The user feedback pointed to a clear pain point, but I needed to analyse and evaluate the main problems with the first iteration of templates, so the next iteration would not repeat some mistakes that might have been done in the past.
The first thing I did was collect data to understand which parts of the template setup needed the most attention, and this is what I discovered:
The main problem in the activation funnel was that only 8% of the users that visited automations started setting up a workflow from templates, while 58% of those users who clicked on a template finished setting it up. This demonstrated that there was an issue in discoverability and users were potentially not aware of how the templates could help them get started.
Discovery
The user feedback pointed to a clear pain point, but I needed to analyse and evaluate the main problems with the first iteration of templates, so the next iteration would not repeat some mistakes that might have been done in the past.
The first thing I did was collect data to understand which parts of the template setup needed the most attention, and this is what I discovered:
The main problem in the activation funnel was that only 8% of the users that visited automations started setting up a workflow from templates, while 58% of those users who clicked on a template finished setting it up. This demonstrated that there was an issue in discoverability and users were potentially not aware of how the templates could help them get started.
Discovery
The user feedback pointed to a clear pain point, but I needed to analyse and evaluate the main problems with the first iteration of templates, so the next iteration would not repeat some mistakes that might have been done in the past.
The first thing I did was collect data to understand which parts of the template setup needed the most attention, and this is what I discovered:
The main problem in the activation funnel was that only 8% of the users that visited automations started setting up a workflow from templates, while 58% of those users who clicked on a template finished setting it up. This demonstrated that there was an issue in discoverability and users were potentially not aware of how the templates could help them get started.
Hypotheses
After diagnosing the main navigation issues, I had a few ideas on why the discoverability of the templates was so low.
Cognitive overload
Too many similar templates, which could make it harder for users to find the right template
Mental model
Titles and descriptions relied heavily on technicalities, which could cause confusion in users who aren't yet familiarized with the feature.
Illustrations
Illustrations were very redundant and repetitive.
One big win
What was spot on, was that the wizard inputs were in natural language, this helped abstract some of the complexity from the workflows and allowed users to grasp what that automation would accomplish.
Hypotheses
After diagnosing the main navigation issues, I had a few ideas on why the discoverability of the templates was so low.
Cognitive overload
Too many similar templates, which could make it harder for users to find the right template
Mental model
Titles and descriptions relied heavily on technicalities, which could cause confusion in users who aren't yet familiarized with the feature.
Illustrations
Illustrations were very redundant and repetitive.
One big win
What was spot on, was that the wizard inputs were in natural language, this helped abstract some of the complexity from the workflows and allowed users to grasp what that automation would accomplish.
Hypotheses
After diagnosing the main navigation issues, I had a few ideas on why the discoverability of the templates was so low.
Cognitive overload
Too many similar templates, which could make it harder for users to find the right template
Mental model
Titles and descriptions relied heavily on technicalities, which could cause confusion in users who aren't yet familiarized with the feature.
Illustrations
Illustrations were very redundant and repetitive.
One big win
What was spot on, was that the wizard inputs were in natural language, this helped abstract some of the complexity from the workflows and allowed users to grasp what that automation would accomplish.
Validation
In effort to understand the current UX and take those learnings into consideration when thinking of the design, I conducted some testing sessions and prompted users with a scenario, they were asked to choose between two templates that were essentially the same, but with different illustrations and titles, to see if they would prefer options that described use cases or detailed steps of the workflow.
Variation A
Variation B
Most of the users were drawn to templates that communicated the use case instead of detailed steps, and the illustrations had no impact in the choice they made.
After discovering that the messaging on the templates should be less descriptive, I gave users a list of templates and asked them to sort it in a way that made sense for them. People would either divide them by action performed or by use case.
Validation
In effort to understand the current UX and take those learnings into consideration when thinking of the design, I conducted some testing sessions and prompted users with a scenario, they were asked to choose between two templates that were essentially the same, but with different illustrations and titles, to see if they would prefer options that described use cases or detailed steps of the workflow.
Variation A
Variation B
Most of the users were drawn to templates that communicated the use case instead of detailed steps, and the illustrations had no impact in the choice they made.
After discovering that the messaging on the templates should be less descriptive, I gave users a list of templates and asked them to sort it in a way that made sense for them. People would either divide them by action performed or by use case.
Validation
In effort to understand the current UX and take those learnings into consideration when thinking of the design, I conducted some testing sessions and prompted users with a scenario, they were asked to choose between two templates that were essentially the same, but with different illustrations and titles, to see if they would prefer options that described use cases or detailed steps of the workflow.
Variation A
Variation B
Most of the users were drawn to templates that communicated the use case instead of detailed steps, and the illustrations had no impact in the choice they made.
After discovering that the messaging on the templates should be less descriptive, I gave users a list of templates and asked them to sort it in a way that made sense for them. People would either divide them by action performed or by use case.
Solution
Card update
Template cards would now be more value driven: Users would be able to quickly spot what that automation does and how it can help them - Illustrations were replaced by icons that describe the steps they need to fill in
Grouped actions
To reduce the clutter in the page, I decided to group templates by type of outcome, so now users can choose which action they would like to perform only after selecting the template.
Filtering and collections
The filters and collections were made to address the 2 different mental models detected in the card sorting sessions - templates grouped by use case or action.
Solution
Card update
Template cards would now be more value driven: Users would be able to quickly spot what that automation does and how it can help them - Illustrations were replaced by icons that describe the steps they need to fill in
Grouped actions
To reduce the clutter in the page, I decided to group templates by type of outcome, so now users can choose which action they would like to perform only after selecting the template.
Filtering and collections
The filters and collections were made to address the 2 different mental models detected in the card sorting sessions - templates grouped by use case or action.
Solution
Card update
Template cards would now be more value driven: Users would be able to quickly spot what that automation does and how it can help them - Illustrations were replaced by icons that describe the steps they need to fill in
Grouped actions
To reduce the clutter in the page, I decided to group templates by type of outcome, so now users can choose which action they would like to perform only after selecting the template.
Filtering and collections
The filters and collections were made to address the 2 different mental models detected in the card sorting sessions - templates grouped by use case or action.
Results
285%
increase on template card clicks
1.5x
bump on automations created from templates
Results
285%
increase on template card clicks
1.5x
bump on automations created from templates
Results
285%
increase on template card clicks
1.5x
bump on automations created from templates
Learnings
There is always a way of validating assumptions, especially when they are hard to back with data.
As this project involved a lot of copy, I had to design things not only thinking of visual aspects, but also how the copy would translate into 22 different languages, and how the design, labels and inputs, would be scalable for localisation purposes.
Learnings
There is always a way of validating assumptions, especially when they are hard to back with data.
As this project involved a lot of copy, I had to design things not only thinking of visual aspects, but also how the copy would translate into 22 different languages, and how the design, labels and inputs, would be scalable for localisation purposes.
Learnings
There is always a way of validating assumptions, especially when they are hard to back with data.
As this project involved a lot of copy, I had to design things not only thinking of visual aspects, but also how the copy would translate into 22 different languages, and how the design, labels and inputs, would be scalable for localisation purposes.