Envato Elements
Envato Elements
Envato Elements
Search and Discovery
Search and Discovery
Search and Discovery
Overview
Envato Elements (or Elements for short) is an online creative subscription, offering people unlimited downloads of millions of digital assets.
For Elements users, the ability to find the right assets on Elements is crucial for them to complete their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business. In 2020, we set out to improve the Elements search functionality and experience and in 2021, we have also set out to make inspirational content more discoverable on Elements.
Here are the outcomes (August - November 2020):
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- 28% increase in search to subscribe rate
- A lower percentage of searches receive no results (8% down to 2%)
Overview
Envato Elements (or Elements for short) is an online creative subscription, offering people unlimited downloads of millions of digital assets.
For Elements users, the ability to find the right assets on Elements is crucial for them to complete their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business. In 2020, we set out to improve the Elements search functionality and experience and in 2021, we have also set out to make inspirational content more discoverable on Elements.
Here are the outcomes (August - November 2020):
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- A lower percentage of searches receive no results (8% down to 2%)
Overview
Envato Elements (or Elements for short) is an online creative subscription, offering people unlimited downloads of millions of digital assets.
For Elements users, the ability to find the right assets on Elements is crucial for them to complete their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business. In 2020, we set out to improve the Elements search functionality and experience and in 2021, we have also set out to make inspirational content more discoverable on Elements.
Here are the outcomes (August - November 2020):
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- A lower percentage of searches receive no results (8% down to 2%)
Overview
Envato Elements (or Elements for short) is an online creative subscription, offering people unlimited downloads of millions of digital assets.
For Elements users, the ability to find the right assets on Elements is crucial for them to complete their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business. In 2020, we set out to improve the Elements search functionality and experience and in 2021, we have also set out to make inspirational content more discoverable on Elements.
Here are the outcomes (August - November 2020):
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- A lower percentage of searches receive no results (8% down to 2%)
My Role
The lead Product Designer for this initiative, working with the core Search and Discovery delivery team consisting of Engineers, Product Managers, a Product Analyst and a UX Researcher.
I also included other subject matter experts and stakeholders in the Content, SEO, Tools and Integration and Marketing space in my process, ensuring that they were part of relevant conversations, relevant feedback loops, joined in workshops I facilitated and understood the rationale behind decisions.
What I've done: UX research, benchmarking, competitor analysis, workshop planning and facilitation (discover, define and ideate), prototyping, concept testing, UX and UI Design, Interaction Design.
Year
2020 - Present
My Role
The lead Product Designer for this initiative, working with the core Search and Discovery delivery team consisting of Engineers, Product Managers, a Product Analyst and a UX Researcher.
I also included other subject matter experts and stakeholders in the Content, SEO, Tools and Integration and Marketing space in my process, ensuring that they were part of relevant conversations, relevant feedback loops, joined in workshops I facilitated and understood the rationale behind decisions.
What I've done: UX research, benchmarking, competitor analysis, workshop planning and facilitation (discover, define and ideate), prototyping, concept testing, UX and UI Design, Interaction Design.
Duration
2020 - Present
My Role
The lead Product Designer for this initiative, working with the core Search and Discovery delivery team consisting of Engineers, Product Managers, a Product Analyst and a UX Researcher.
I also included other subject matter experts and stakeholders in the Content, SEO, Tools and Integration and Marketing space in my process, ensuring that they were part of relevant conversations, relevant feedback loops, joined in workshops I facilitated and understood the rationale behind decisions.
What I've done: UX research, benchmarking, competitor analysis, workshop planning and facilitation (discover, define and ideate), prototyping, concept testing, UX and UI Design, Interaction Design.
Duration
2020 - Present
My Role
The lead Product Designer for this initiative, working with the core Search and Discovery delivery team consisting of Engineers, Product Managers, a Product Analyst and a UX Researcher.
I also included other subject matter experts and stakeholders in the Content, SEO, Tools and Integration and Marketing space in my process, ensuring that they were part of relevant conversations, relevant feedback loops, joined in workshops I facilitated and understood the rationale behind decisions.
What I've done: UX research, benchmarking, competitor analysis, workshop planning and facilitation (discover, define and ideate), prototyping, concept testing, UX and UI Design, Interaction Design.
Duration
2020 - Present
My Role
The lead Product Designer for this initiative, working with the core Search and Discovery delivery team consisting of Engineers, Product Managers, a Product Analyst and a UX Researcher.
I also included other subject matter experts and stakeholders in the Content, SEO, Tools and Integration and Marketing space in my process, ensuring that they were part of relevant conversations, relevant feedback loops, joined in workshops I facilitated and understood the rationale behind decisions.
What I've done: UX research, benchmarking, competitor analysis, workshop planning and facilitation (discover, define and ideate), prototyping, concept testing, UX and UI Design, Interaction Design.
Duration
2020 - Present
Starting with Search
Discover
For Elements users (potential and existing subscribers), the ability to search on Elements is crucial in finding the right assets for their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business because:
- Search is the most frequently mentioned reason for user dissatisfaction with Elements (CSAT).
- Search is used in >75% of subscriber sessions.
- Improving search will lead to an increase in revenue and community earnings through more downloads, subscriptions as well as improved customer lifetime value.
- Market research has shown that Search & Discovery is heavily invested in by competitors.
- With Element’s growing library size, search becomes more important to enable our users to find what they need quickly.
“It takes HOURS to dig for something that takes a few seconds to search in any other stock site...”
- Elements user
Examples of insights and data
Qualitative and Quantitative insights (gathered from user interviews and surveying):
- Finding relevant assets efficiently is key for users and an accurate keyword search is extremely important in finding relevant assets along with different filtering options.
- Taking too long to search for assets is a major pain point as many Elements users are time-poor. If it feels like a search is taking too long, users will look to other providers.
- Perceived lack of variety and diversity in content is a driver of dissatisfaction when searching for assets.
- Many users would like personalisation features to enable more spontaneous content discovery
Note: We spoke with 12 participants, including 5 potential users and 7 current subscribers.
Analysis of Elements visitor behaviour when searching (measured over 3 weeks at the time):
- 41% of visitors searched on-site
- 51% of people that searched used one keyword and 49% used 2 or more keywords
- In 94% of search queries, visitors only look at the first 4 pages before exiting
- 75% of clicks on an item in search results happen on the first page
- 8% of searches come back up with 0 results.
Competitor analysis:
- Competitors provide more ways for users to refine their search e.g. more filtering options and more metadata to work with
- More curated and trending content is found on competitor sites.
Other insights gathered when speaking with subject matter experts and doing internal reviews:
- We need to better utilise the metadata we capture from authors (creators of our assets) and improve the quality of metadata associated with each asset to enable us to surface more relevant and accurate search results.
- There are existing issues on-site e.g. Search picks up on partial matches AND search picking up exact matches, Search ignoring conjunction words like: and, but
- We should determine whether the search algorithm should change per item type.
- In the future, we should look into machine learning capabilities to respond to what our users need.
My role involved assisting with user research, synthesis, leading the competitor analysis and discussions with subject matter experts.
Starting with Search
Discover
For Elements users (potential and existing subscribers), the ability to search on Elements is crucial in finding the right assets for their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business because:
- Search is the most frequently mentioned reason for user dissatisfaction with Elements (CSAT).
- Search is used in >75% of subscriber sessions.
- Improving search will lead to an increase in revenue and community earnings through more downloads, subscriptions as well as improved customer lifetime value.
- Market research has shown that Search & Discovery is heavily invested in by competitors.
- With Element’s growing library size, search becomes more important to enable our users to find what they need quickly.
“It takes HOURS to dig for something that takes a few seconds to search in any other stock site...”
- Elements user
Examples of insights and data
Qualitative and Quantitative insights (gathered from user interviews and surveying):
- Finding relevant assets efficiently is key for users and an accurate keyword search is extremely important in finding relevant assets along with different filtering options.
- Taking too long to search for assets is a major pain point as many Elements users are time-poor. If it feels like a search is taking too long, users will look to other providers.
- Perceived lack of variety and diversity in content is a driver of dissatisfaction when searching for assets.
- Many users would like personalisation features to enable more spontaneous content discovery
Note: We spoke with 12 participants, including 5 potential users and 7 current subscribers.
Analysis of Elements visitor behaviour when searching (measured over 3 weeks at the time):
- 41% of visitors searched on-site
- 51% of people that searched used one keyword and 49% used 2 or more keywords
- In 94% of search queries, visitors only look at the first 4 pages before exiting
- 75% of clicks on an item in search results happen on the first page
- 8% of searches come back up with 0 results.
Competitor analysis:
- Competitors provide more ways for users to refine their search e.g. more filtering options and more metadata to work with
- More curated and trending content is found on competitor sites.
Other insights gathered when speaking with subject matter experts and doing internal reviews:
- We need to better utilise the metadata we capture from authors (creators of our assets) and improve the quality of metadata associated with each asset to enable us to surface more relevant and accurate search results.
- There are existing issues on-site e.g. Search picks up on partial matches AND search picking up exact matches, Search ignoring conjunction words like: and, but
- We should determine whether the search algorithm should change per item type.
- In the future, we should look into machine learning capabilities to respond to what our users need.
My role involved assisting with user research, synthesis, leading the competitor analysis and discussions with subject matter experts.
Starting with Search
Discover
For Elements users (potential and existing subscribers), the ability to search on Elements is crucial in finding the right assets for their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business because:
- Search is the most frequently mentioned reason for user dissatisfaction with Elements (CSAT).
- Search is used in >75% of subscriber sessions.
- Improving search will lead to an increase in revenue and community earnings through more downloads, subscriptions as well as improved customer lifetime value.
- Market research has shown that Search & Discovery is heavily invested in by competitors.
- With Element’s growing library size, search becomes more important to enable our users to find what they need quickly.
“It takes HOURS to dig for something that takes a few seconds to search in any other stock site...”
- Elements user
Examples of insights and data
Qualitative and Quantitative insights (gathered from user interviews, surveys and internal reviews):
- Finding relevant assets efficiently is key for users and an accurate keyword search is extremely important in finding relevant assets along with different filtering options.
- Taking too long to search for assets is a major pain point as many Elements users are time-poor. If it feels like a search is taking too long, users will look to other providers.
- Perceived lack of variety and diversity in content is a driver of dissatisfaction when searching for assets.
- Many users would like personalisation features to enable more spontaneous content discovery
Note: We spoke with 12 participants, including 5 potential users and 7 current subscribers.
Analysis of Elements visitor behaviour when searching (measured over 3 weeks at the time):
- 41% of visitors searched on-site
- 51% of people that searched used one keyword and 49% used 2 or more keywords
- In 94% of search queries, visitors only look at the first 4 pages before exiting
- 75% of clicks on an item in search results happen on the first page
- 8% of searches come back up with 0 results.
Competitor analysis:
- Competitors provide more ways for users to refine their search e.g. more filtering options and more metadata to work with
- More curated and trending content is found on competitor sites.
Other insights gathered when speaking with subject matter experts and doing internal reviews:
- We need to better utilise the metadata we capture from authors (creators of our assets) and improve the quality of metadata associated with each asset to enable us to surface more relevant and accurate search results.
- There are existing issues on-site e.g. Search picks up on partial matches AND search picking up exact matches, Search ignoring conjunction words like: and, but
- We should determine whether the search algorithm should change per item type.
- In the future, we should look into machine learning capabilities to respond to what our users need.
My role involved assisting with user research, synthesis, leading the competitor analysis and discussions with subject matter experts.
Starting with Search
Discover
For Elements users (potential and existing subscribers), the ability to search on Elements is crucial in finding the right assets for their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business because:
- Search is the most frequently mentioned reason for user dissatisfaction with Elements (CSAT).
- Search is used in >75% of subscriber sessions.
- Improving search will lead to an increase in revenue and community earnings through more downloads, subscriptions as well as improved customer lifetime value.
- Market research has shown that Search & Discovery is heavily invested in by competitors.
- With Element’s growing library size, search becomes more important to enable our users to find what they need quickly.
“It takes HOURS to dig for something that takes a few seconds to search in any other stock site...”
- Elements user
Examples of insights and data
Qualitative and Quantitative insights (gathered from user interviews, surveys and internal reviews):
- Finding relevant assets efficiently is key for users and an accurate keyword search is extremely important in finding relevant assets along with different filtering options.
- Taking too long to search for assets is a major pain point as many Elements users are time-poor. If it feels like a search is taking too long, users will look to other providers.
- Perceived lack of variety and diversity in content is a driver of dissatisfaction when searching for assets.
- Many users would like personalisation features to enable more spontaneous content discovery
Note: We spoke with 12 participants, including 5 potential users and 7 current subscribers.
Analysis of Elements visitor behaviour when searching (measured over 3 weeks at the time):
- 41% of visitors searched on-site
- 51% of people that searched used one keyword and 49% used 2 or more keywords
- In 94% of search queries, visitors only look at the first 4 pages before exiting
- 75% of clicks on an item in search results happen on the first page
- 8% of searches come back up with 0 results.
Competitor analysis:
- Competitors provide more ways for users to refine their search e.g. more filtering options and more metadata to work with
- More curated and trending content is found on competitor sites.
Other insights gathered when speaking with subject matter experts and doing internal reviews:
- We need to better utilise the metadata we capture from authors (creators of our assets) and improve the quality of metadata associated with each asset to enable us to surface more relevant and accurate search results.
- There are existing issues on-site e.g. Search picks up on partial matches AND search picking up exact matches, Search ignoring conjunction words like: and, but
- We should determine whether the search algorithm should change per item type.
- In the future, we should look into machine learning capabilities to respond to what our users need.
My role involved assisting with user research, synthesis, leading the competitor analysis and discussions with subject matter experts.
Starting with Search
Discover
For Elements users (potential and existing subscribers), the ability to search on Elements is crucial in finding the right assets for their creative projects. However, the search experience on Elements is less than ideal and a key improvement area for the business because:
- Search is the most frequently mentioned reason for user dissatisfaction with Elements (CSAT).
- Search is used in >75% of subscriber sessions.
- Improving search will lead to an increase in revenue and community earnings through more downloads, subscriptions as well as improved customer lifetime value.
- Market research has shown that Search & Discovery is heavily invested in by competitors.
- With Element’s growing library size, search becomes more important to enable our users to find what they need quickly.
“It takes HOURS to dig for something that takes a few seconds to search in any other stock site...”
- Elements user
Examples of insights and data
Qualitative and Quantitative insights (gathered from user interviews and surveying):
- Finding relevant assets efficiently is key for users and an accurate keyword search is extremely important in finding relevant assets along with different filtering options.
- Taking too long to search for assets is a major pain point as many Elements users are time-poor. If it feels like a search is taking too long, users will look to other providers.
- Perceived lack of variety and diversity in content is a driver of dissatisfaction when searching for assets.
- Many users would like personalisation features to enable more spontaneous content discovery
Note: We spoke with 12 participants, including 5 potential users and 7 current subscribers.
Analysis of Elements visitor behaviour when searching (measured over 3 weeks at the time):
- 41% of visitors searched on-site
- 51% of people that searched used one keyword and 49% used 2 or more keywords
- In 94% of search queries, visitors only look at the first 4 pages before exiting
- 75% of clicks on an item in search results happen on the first page
- 8% of searches come back up with 0 results.
Competitor analysis:
- Competitors provide more ways for users to refine their search e.g. more filtering options and more metadata to work with
- More curated and trending content is found on competitor sites.
Other insights gathered when speaking with subject matter experts and doing internal reviews:
- We need to better utilise the metadata we capture from authors (creators of our assets) and improve the quality of metadata associated with each asset to enable us to surface more relevant and accurate search results.
- There are existing issues on-site e.g. Search picks up on partial matches AND search picking up exact matches, Search ignoring conjunction words like: and, but
- We should determine whether the search algorithm should change per item type.
- In the future, we should look into machine learning capabilities to respond to what our users need.
My role involved assisting with user research, synthesis, leading the competitor analysis and discussions with subject matter experts.
Describe
Our core search team ran a kick-off workshop with stakeholders and subject matter experts to look at the problem space, align on why we’re addressing this problem, list out questions, assumptions and agree on the outcomes we wanted.
Desired outcomes (details not shared):
- Increase downloads
- Increase subscriptions from search
- Increase user satisfaction
- Increase interactions from searches
- Decrease exit rate from search pages
Following that, we ran an opportunity tree creation session to come up with a set of user statements (informed by the insights and data) under the outcomes and came up with opportunities and potential solutions under the statements. We then chose 2 statements to prioritise going forward based on potential impact, value and effort:
- As an Elements user, I don't get relevant results on the first page of my search. I’m unsure if you don’t have the assets or if I’m not entering the right keyword/s.
- As an Elements user, I want to be able to save time and find assets as efficiently as possible.
Opportunity tree
I also facilitated an ideation workshop with the core search team to generate and encourage the exploration of more ideas around how we can help customers find assets faster (statement 2) through the reverse brainstorming approach.
Screenshot of people's sketches in the ideation workshop - How might we allow users to save time and find assets as efficiently as possible?
Screenshot of people's sketches in the ideation workshop - How might we allow users to save time and find assets as efficiently as possible?
Design, Develop
Rapid experimentation
We first made improvements and fixes to the basic search functionality e.g. better handling of compound words and synonyms and how we handle stemming (e.g. words ending in “-ing”) to serve more relevant results to users.
In order to get to a greater degree of confidence in which solutions would be best in addressing the 2 user statements, we did on-site multivariate testing and A/B testing in sprints (tests ran for 2 weeks or more on-site, determined by the amount of traffic needed) to validate the product, UX and algorithm related solutions we came up with. At the end of each test, a decision was made on whether we would deploy, iterate, pivot, shelve or abandon the solution based on whether it was helping us achieve our desired outcomes or not.
We had a hypothesis (informed by existing insights and data) for each solution we tested and determined what we were tracking to measure impact. Having a hypothesis is important to avoid testing just for the sake of it.
For product, UX and design-related opportunities, I formed and shaped hypotheses with the input of the core team and subject matter experts and explored different designs of variants we could test. During this stage, I like to get people from other teams e.g. digital designers and product designers from other parts of the business involved in co-design sessions whenever I can to get fresh perspectives and come up with more designs. I also did concept testing to gather user feedback and had constant feedback loops to guide the direction of designs and help me rationalise and present design decisions back to my team and stakeholders.
Design solution example: Genre, Mood and Instruments
Hypothesis: Customer feedback from as early as 2018 gave insight that Elements audio users want the ability to refine their audio items by Genre, Mood and Instrument (GMI). We hypothesise that it will help users find relevant audio items more efficiently and in turn, increase downloads and conversion.
What we tracked: Usage, interaction with audio items, search to download rate, search to subscribe rate.
I worked closely with audio content specialists to synthesise customer feedback, do a competitor analysis and identify genres, moods and instruments among the top search keywords. This helped me create separate lists of Genre, Mood and Instrument terms to test as facets. The list was validated against our data and can be refined as we learn more about Audio user behaviour and needs over time.
Design, Develop
Rapid experimentation
We first made improvements and fixes to the basic search functionality e.g. better handling of compound words and synonyms and how we handle stemming (e.g. words ending in “-ing”) to serve more relevant results to users.
In order to get to a greater degree of confidence in which solutions would be best in addressing the 2 user statements, we did on-site multivariate testing and A/B testing in sprints (tests ran for 2 weeks or more on-site, determined by the amount of traffic needed) to validate the product, UX and algorithm related solutions we came up with. At the end of each test, a decision was made on whether we would deploy, iterate, pivot, shelve or abandon the solution based on whether it was helping us achieve our desired outcomes or not.
We had a hypothesis (informed by existing insights and data) for each solution we tested and determined what we were tracking to measure impact. Having a hypothesis is important to avoid testing just for the sake of it.
For product, UX and design-related opportunities, I formed and shaped hypotheses with the input of the core team and subject matter experts and explored different designs of variants we could test. During this stage, I like to get people from other teams e.g. digital designers and product designers from other parts of the business involved in co-design sessions whenever I can to get fresh perspectives and come up with more designs. I also did concept testing to gather user feedback and had constant feedback loops to guide the direction of designs and help me rationalise and present design decisions back to my team and stakeholders.
Design solution example: Genre, Mood and Instruments
Hypothesis: Customer feedback from as early as 2018 gave insight that Elements audio users want the ability to refine their audio items by Genre, Mood and Instrument (GMI). We hypothesise that it will help users find relevant audio items more efficiently and in turn, increase downloads and conversion.
What we tracked: Usage, interaction with audio items, search to download rate, search to subscribe rate.
I worked closely with audio content specialists to synthesise customer feedback, do a competitor analysis and identify genres, moods and instruments among the top search keywords. This helped me create separate lists of Genre, Mood and Instrument terms to test as facets. The list was validated against our data and can be refined as we learn more about Audio user behaviour and needs over time.
After, I explored (using low-fidelity sketches and mock-ups) how users could interact with Genre, Mood and Instrument facets and how it could be displayed on-site for feedback. I created prototypes to run concept testing sessions with 8 Elements audio users to understand how they could utilise genre, mood and instrument facets to find the audio items they need for their projects:
- Do they expect it to be part of the existing filters?
- Will the ability to exclude as well as include genres, moods and instruments be valuable?
- In what way should genre, mood and instrument facets be presented that will make it easy for audio users to find what they need?
Note: The prototype was for desktop as the majority visit and use Elements on desktop.
After, I explored (using low-fidelity sketches and mock-ups) how users could interact with Genre, Mood and Instrument facets and how it could be displayed on-site for feedback. I created prototypes to run concept testing sessions with 8 Elements audio users to understand how they could utilise genre, mood and instrument facets to find the audio items they need for their projects:
- Do they expect it to be part of the existing filters?
- Will the ability to exclude as well as include genres, moods and instruments be valuable?
- In what way should genre, mood and instrument facets be presented that will make it easy for audio users to find what they need?
Note: The prototypes were for desktop as the majority visit and use Elements on desktop.
Screenshot of prototype 1: GMI in the existing filters section on the left
Screenshot of prototype 2: GMI above results
Screenshot of prototype 2: GMI above results
Screenshot of prototype 3: GMI below header
Concept testing GMI with Elements Audio users
Concept testing GMI with Elements Audio users
What we learnt from concept testing:
- While a few users liked having Genre Mood and Instrument as a section above the results (prototype 2) as it was easy to discover, the majority preferred to have it as part of the existing filters on the left (prototype 1) as it’s where they expect it to be (intuitive) and they found it easier to read down the list of facets.
- Most participants find the ability to exclude as well as include GMI useful.
Therefore, this is the direction I headed in for the high-fidelity designs and what we tested against the control (no GMI) on-site.
Results of the A/B test:
- 5.4% increase in overall site conversion
- 3.6% increase in audio downloads
- 4.8% increase in audio play rate
GMI has since been deployed, check it out live here.
Examples of other design solutions (tested and deployed):
Sticky filters
Originally the filters on search pages moved up with the page as the user scrolled down to view the results.
Original filters that were not sticky (control)
Hypothesis: Making the filters sticky allows them to be more accessible for users as they scroll down the page. Users will be able to refine their results to find something relevant more quickly, which will in turn increase downloads and subscriptions.
Outcomes:
- 8% increase in search to subscribe rate (+2% increase in main site conversion)
- No change to search to download rate
Sticky filters have since been deployed. Check it out here.
High-fidelity sticky filter designs (mobile and tablet)
High-fidelity sticky filter designs (desktop)
Related searches
Hypothesis: Related searches are violet links at the top of search pages and users have given us feedback that the feature has been mistaken as breadcrumbs, causing confusion. We hypothesise that:
- If we improved the look & feel of 'related searches', it will reduce user confusion, increase engagement & result in more downloads.
- If we find the optimal position for the ‘related searches’ on results pages, it will better align with the user's search journey and needs.
The original violet related search links that people mistook as breadcrumb navigation (control)
Related searches design exploration: Rectangular pill (search term + thumbnail)
Related searches design exploration: Search term only
Related searches design exploration: Circular pill (search term + thumbnail)
Related searches design exploration: Square thumbnail pill (search term + thumbnail)
Related searches design exploration: Large thumbnail (middle of results)
Related searches design exploration: Large thumbnail with overlaying search term (middle of results)
Variant 3 (Circular pill, at top and bottom of the results) was the winner.
- Out of the visitors who interacted with the winning variant, the subscription rate increased by more than 5%
- The percentage of searches that lead to an exit has decreased for people interacting with the winning variant by 53%.
The new related searches design and position has been deployed, check it out here.
Variant 3 (Circular pill, at top and bottom of the results) was the winner.
- Out of the visitors who interacted with the winning variant, the subscription rate increased by more than 5%
- The percentage of searches that lead to an exit has decreased for people interacting with the winning variant by 53%.
The new related searches design and position has been deployed, check it out here.
Variant 3 (Circular pill, at top and bottom of the results) was the winner.
- Out of the visitors who interacted with the winning variant, the subscription rate increased by more than 5%
- The percentage of searches that lead to an exit has decreased for people interacting with the winning variant by 53%.
The new related searches design and position has been deployed, check it out here.
Variant 3 (Circular pill, at top and bottom of the results) was the winner.
- Out of the visitors who interacted with the winning variant, the subscription rate increased by more than 5%
- The percentage of searches that lead to an exit has decreased for people interacting with the winning variant by 53%.
The new related searches design and position has been deployed, check it out here.
Variant 3 (Circular pill, at top and bottom of the results) was the winner.
- Out of the visitors who interacted with the winning variant, the subscription rate increased by more than 5%
- The percentage of searches that lead to an exit has decreased for people interacting with the winning variant by 53%.
The new related searches design and position has been deployed, check it out here.
Deploy
Here are examples of some of the solutions we’ve deployed so far to improve search and the statements they address
As an Elements user, I don't get relevant results on the first page of my search. I’m unsure if you don’t have the asset or if I’m not entering the right keyword/s:
- Weighting changes to the algorithm
- Fixed exact/partial matching
- User signals
- Compound words
As an Elements user, I want to be able to save time and find assets as efficiently as possible.
- Auto-suggest
- Genre, Mood and Instrument facets in audio
- Sticky filters
- More items on a page
- Improved related searches
- Search by author
- Search within an author’s portfolio
After our first series of changes made to search, I launched a survey to help measure how customers felt with these changes and learn more about how we can continue to improve search. This survey will continue to be launched so that we continuously gather customer feedback and sentiment.
Outcomes so far (August-November 2020)
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- Lower percentage of searches receive no results (8% down to 2%)
Next steps
As finding relevant items remain a core need of our users, we continue investing in:
- High quality metadata to enable more relevant results to be surfaced and more valuable refinements to be made by users.
- Machine learning and other means of increasing the quality of the results.
- Research to dig deeper into the categories and get a better understanding of the content needs of power users for each category.
Deploy
Here are examples of some of the solutions we’ve deployed so far to improve search and the statements they address
As an Elements user, I don't get relevant results on the first page of my search. I’m unsure if you don’t have the asset or if I’m not entering the right keyword/s:
- Weighting changes to the algorithm
- Fixed exact/partial matching
- User signals
- Compound words
As an Elements user, I want to be able to save time and find assets as efficiently as possible.
- Auto-suggest
- Genre, Mood and Instrument facets in audio
- Sticky filters
- More items on a page
- Improved related searches
- Search by author
- Search within an author’s portfolio
After our first series of changes made to search, I launched a survey to help measure how customers felt with these changes and learn more about how we can continue to improve search. This survey will continue to be launched so that we continuously gather customer feedback and sentiment.
Outcomes so far (August-November 2020)
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- Lower percentage of searches receive no results (8% down to 2%)
Next steps
As finding relevant items remain a core need of our users, we continue investing in:
High quality metadata to enable more relevant results to be surfaced and more valuable refinements to be made by users.
Machine learning and other means of increasing the quality of the results.
Research to dig deeper into the categories and get a better understanding of the content needs of power users for each category.
Deploy
Here are examples of some of the solutions we’ve deployed so far to improve search and the statements they address
As an Elements user, I don't get relevant results on the first page of my search. I’m unsure if you don’t have the asset or if I’m not entering the right keyword/s:
- Weighting changes to the algorithm
- Fixed exact/partial matching
- User signals
- Compound words
As an Elements user, I want to be able to save time and find assets as efficiently as possible.
- Auto-suggest
- Genre, Mood and Instrument facets in audio
- Sticky filters
- More items on a page
- Improved related searches
- Search by author
- Search within an author’s portfolio
After our first series of changes made to search, I launched a survey to help measure how customers felt with these changes and learn more about how we can continue to improve search. This survey will continue to be launched so that we continuously gather customer feedback and sentiment.
Outcomes so far (August-November 2020)
- 3% increase in user satisfaction (how easy it is to find what you are looking for through search functionality)
- 19% increase in the search to download rate
- Over 29% increase in search to subscribe rate
- Lower percentage of searches receive no results (8% down to 2%)
Next steps
As finding relevant items remain a core need of our users, we continue investing in:
- High quality metadata to enable more relevant results to be surfaced and more valuable refinements to be made by users.
- Machine learning and other means of increasing the quality of the results.
- Research to dig deeper into the categories and get a better understanding of the content needs of power users for each category.
Diving into the Discovery space (while continuing to improve search)
The beginning
For Elements users, finding inspiration is a key part of their creative process. 80% of Elements users hope to be inspired and for 20%, it’s the hardest part of the creative process. Inspiration needs arise early and all throughout the process, which makes the discovery of inspirational content on Elements a great opener for attracting & converting new users to and retaining existing subscribers.
To kick things off in the discovery space, I planned and facilitated a 3-day design sprint with the core search delivery team, stakeholders across the business and users (both in-person and remote).
The overall objective of the design sprint: Identify the goal we wish to achieve by improving Discovery on Elements, the key customer problems we are solving and an idea of the solutions we would explore
Diving into the Discovery space (while continuing to improve search)
The beginning
For Elements users, finding inspiration is a key part of their creative process. 80% of Elements users hope to be inspired and for 20%, it’s the hardest part of the creative process. Inspiration needs arise early and all throughout the process, which makes the discovery of inspirational content on Elements a great opener for attracting & converting new users to and retaining existing subscribers.
To kick things off in the discovery space, I planned and facilitated a 3-day design sprint with the core search delivery team, stakeholders across the business and users (both in-person and remote).
The overall objective of the design sprint: Identify the goal we wish to achieve by improving Discovery on Elements, the key customer problems we are solving and an idea of the solutions we would explore.
Diving into the Discovery space (while continuing to improve search)
The beginning
For Elements users, finding inspiration is a key part of their creative process. 80% of Elements users hope to be inspired and for 20%, it’s the hardest part of the creative process. Inspiration needs arise early and all throughout the process, which makes the discovery of inspirational content on Elements a great opener for attracting & converting new users to and retaining existing subscribers.
To kick things off in the discovery space, I planned and facilitated a 3-day design sprint with the core search delivery team, stakeholders across the business and users (both in-person and remote).
The overall objective of the design sprint: Identify the goal we wish to achieve by improving Discovery on Elements, the key customer problems we are solving and an idea of the solutions we would explore.
Diving into the Discovery space (while continuing to improve search)
The beginning
For Elements users, finding inspiration is a key part of their creative process. 80% of Elements users hope to be inspired and for 20%, it’s the hardest part of the creative process. Inspiration needs arise early and all throughout the process, which makes the discovery of inspirational content on Elements a great opener for attracting & converting new users to and retaining existing subscribers.
To kick things off in the discovery space, I planned and facilitated a 3-day design sprint with the core search delivery team, stakeholders across the business and users (both in-person and remote).
The overall objective of the design sprint: Identify the goal we wish to achieve by improving Discovery on Elements, the key customer problems we are solving and an idea of the solutions we would explore.
Diving into the Discovery space (while continuing to improve search)
The beginning
For Elements users, finding inspiration is a key part of their creative process. 80% of Elements users hope to be inspired and for 20%, it’s the hardest part of the creative process. Inspiration needs arise early and all throughout the process, which makes the discovery of inspirational content on Elements a great opener for attracting & converting new users to and retaining existing subscribers.
To kick things off in the discovery space, I planned and facilitated a 3-day design sprint with the core search delivery team, stakeholders across the business and users (both in-person and remote).
The overall objective of the design sprint: Identify the goal we wish to achieve by improving Discovery on Elements, the key customer problems we are solving and an idea of the solutions we would explore.
Images taken from the Discovery Design sprint