OfferUp • April 2023
Simplifying the Transaction Experience - How OfferUp Can Reduce Headaches For its Buyer and Sellers
CONTEXT
Is buying and selling on OfferUp as simple as it could be?
On its website, OfferUp states it is the simplest, most trusted way to buy and sell.
But is this really the case?
Friends who were buying items on OfferUp complained that it took too long for sellers to return their messages. They used OfferUp to simplify their purchasing experience but were not satisfied.
I decided to dive deeper and evaluate how OfferUp can innovate its product to provide an enhanced user experience for buyers and sellers.
If OfferUp implemented more automated features in its messaging functionality, that would expedite the sales process and increase the rate of sales.
ORIGINAL HYPOTHESIS
PRODUCT DISCOVERY INSIGHTS
User interviews were conducted with a mix of both buyers and sellers on the OfferUp platform.
I found that the most critical force driving any transaction is the “ease of process”.
How quickly can product questions get answered?
How low of a lift can we put on sellers to respond to numerous potential buyers in a reasonable amount of time?
Talking to OfferUp Users
90% of interviewed found that the back-and-forth messaging was discouraging to sellers. Buyers haggling, and asking general questions leaves sellers feeling pessimistic about the chances of a sale happening.
60% interviewed noted that seller accountability is of critical importance on the platform. Interviewees gave positive feedback on the rating system, but a more stringent verification process would build greater confidence for buyers.
60% of customers interviewed found that the buyer base was untrustworthy when it came to either negotiating or in-person purchases. Safety was a primary concern among female sellers.
“I would rely on the automated options simply because of the lack of time and the large volume of interest I get.”
“I would follow up and then the seller let me know that the item had been sold but I think much of it could be people aren’t checking the app unless they’re actively using it to sell for their small business.”
User Quotes
“Folks aren’t checking the app and when they do, they respond to whoever demonstrated interest most recently to sell the item.”
AUDIENCE INSIGHTS
Strong Market Growth
According to research by OfferUp, the recommerce (retail e-commerce) market is projected to grow 58% by 2028, reaching $276 billion.
Who uses OfferUp?
Source: OfferUp
The most substantial portion of users are males aged 25-34 and 35-44.
Source: Similarweb
Customer Personas
Bargain Buyer
Target Audience Summary
Gender: Male
Age: 22-34
Income - $0k-$70k
Interests - electronics, cars, sports
Pain points - retail stores too expensive, can’t find affordable options
Motivations - save money, score great deals, maintain desired lifestyle
Persona - cost-conscious bachelor
Use Cases of OfferUp
Finding items that meet similar quality standards to retail listings at a lower price
Easily find items without searching in physical retail stores
Compare products by quality and price to land the best deal
Super Seller
Target Audience Summary
Gender: Male
Age: 35-44
Income - $70k-$200k
Interests - collectibles, minimalism, entrepreneurship, environmentalism
Pain points - Have too much clutter, guilty throwing away items still in good condition
Motivations - earn extra income, declutter home
Persona - money-savvy entrepreneur
Use Cases of OfferUp
Quickly identify potential buyers
Get compensated for items no longer needed
Dispose of possessions without trashing them
How does a buyer communicate with a seller to answer simple questions?
USER JOURNEY
BIG TAKEAWAYS
BIG TAKEAWAYS
From this research, we can conclude a couple of things:
Users are predominately young males seeking bargain deals and getting rid of unnecessary items.
The biggest hassle for sellers is having to answer the same questions repeatedly for multiple buyers.
Many buyers do not read listing descriptions carefully, while many sellers are not able to answer questions in a timely manner, causing the transaction to stall.
THE PROBLEM
Too many unnecessary messages are exchanged between buyers and sellers before transactions occur, and it all happens too slowly.
Build an automated messaging feature for sellers to answer commonly asked questions from buyers.
THE GOAL
Implementing these changes will decrease the average chat time and increase the rate at which transactions are finalized.
What should be included in the MVP?
FEATURE PRIORITIZATION & MVP DEFINITION
There are a variety of pain points for users including messaging, safety, and verification. We believe the messaging component should be the primary pain point addressed considering the value delivered and the effort needed to deliver on this feature.
User Stories
-
Acceptance Criteria: Given the seller has inputted answers to a list of prompts that feature the most frequent questions, the seller can simply turn on automated messages in the settings/details page of the item listing.
Design: Common questions prompts will be featured on the details page prior to posting the item for sale. Toggle button to turn on and off automated messages. Automated messages will appear as lime green text bubbles in a chat. This feature can be altered before or after listing the item.
-
Acceptance Criteria: Given the seller has set the lowest acceptable offer under the price page, OfferUp will automatically reject any buyer offers that fall below the Seller’s limit. A custom message can also be included to be included with the rejection notice.
Design: Sellers will be prompted to set the lowest acceptable offer along with a box to input a custom rejection message on the Price page. This feature can be edited before or after the listing has gone live.
-
Acceptance Criteria: Given the buyer has asked a question to the Seller about the item, the Buyer will instantly receive a response from the Seller that pertains to their question.
Design: Once the platform has identified the Buyer’s text as a commonly asked question, the correct answer will be sent instantly as a chat response. The text bubble will be lime green to infer that the message was automated.
AI-Generated Responses to Instantaneously Answer the Most Commonly Asked Questions
FINAL SOLUTION
Seller turns on Deal Assistant feature to automatically respond to common questions.
When the buyer asks a common question, an automated response is generated.
Note: automated messaged in light green and manual messages in dark green
What risks would OfferUp face by adding the Deal Assistant feature?
RISKS & TRADEOFFS
Sellers not updating automated information
The seller could have changing answers after inputting responses, resulting in the autogenerated answers providing incorrect information for the buyer (meeting time availability, product availability, etc.)
The influx of incorrect information could lead to greater dissatisfaction on the buyer’s part.
Potential mitigation strategy - integrate seller’s calendars (Apple/Google calendar) to update the seller’s pickup availability accordingly.
Buyers expecting every response will be immediate
For simplicity reasons, there will need to be a limited number of automated responses for the initial MVP release.
Buyers may ask a seemingly simple question and become impatient sooner than normal.
Potential mitigation strategy - there can be a disclaimer for messages that aren’t automated responses to set expectations on response times.
Time & resources it’ll take to build, launch, and maintain the Deal Assistant feature
Adding Deal Assistant to the UX adds another layer of complexity to maintaining the OfferUp app. We have to ensure the new feature is delivering a positive user experience.
Committing to the Deal Assistant feature means fewer resources can be dedicated to adding other new app features.
MEASURING SUCCESS
A/B Test Metrics
NORTH STAR METRIC
Reduction of Time between First Buyer Inquiry and Final Sale
If we are solving for the pain point of lengthy messaging by utilizing the Deal Assistant feature, then we should see a reduction of time between when a buyer first asks a question about the product and when the product is sold.
SECONDARY
These secondary metrics will help us determine the effectiveness of the new feature and potential areas for improvement.
Percentage of Sellers who Toggle on Deal Assistant
What percentage of sellers utilize the Deal Assistant feature?
Average Number of Messages Exchanged
What is the average number of messages exchanged between the buyer and seller before a transaction is made?
How does this compare to the average number of messages exchanged for sellers who don’t utilize the Deal Assistant feature?
Percentage of Seller Messages that are Autogenerated
What percentage of the seller messages are autogenerated versus manually inputted?
Percentage of Successful Transactions
What percentage of chats utilizing Deal Assitant result in successful transactions?
How does this rate compare to the success rate of transactions that don’t utilize Deal Assistant?
COUNTER METRICS
Since the Deal Assistant feature will allow sellers to not have to spend as much time viewing and responding to messages, we want to ensure we are not unintentionally hurting other core metrics. To monitor this, we will track:
Refund Rate
Are buyers requesting refunds at a higher rate for products sold by Sellers who have Deal Assistant on?
Support Requests
Are buyers reaching out to Support at higher rates because misinformation or confusion due to Deal Assistant? Do benefits from this feature outweigh the increased Support costs associated with launching this feature?
Lower Average Seller Spend
Are sellers opening up the app less due to automated messages ultimately resulting in less time exploring other products to buy?
Delayed Response Time by Sellers Responding to Manual Messages
Do sellers become overly reliant on automated messages, and therefore check the app less resulting in longer response times?
A/B Test Deal Assistant
LAUNCH & GTM STRATEGY
To ensure we are effectively reducing the amount of time it takes to finalize a sale, we will be conducting an A/B test. The test will involve a small group of existing users within our target user segment.
Versions:
Control: Current OfferUp users without access to the Deal Assistant feature (Start at 90% of audience)
Variant: OfferUp users with access to the Deal Assistant feature (Start at 10% of audience)
Primary Metric: Number of minutes elapsed between initial buyer message and completed sale
We will be starting the A/B test variant at 10% of our target audience so that we can iterate without affecting the majority of users in this audience.
If the results of the A/B test are positive (primary measure of success achieved and no negative side effects) we will roll out Deal Assistant to all users in the United States. Eventually, we would roll out the MVP to all users internationally.
From there, we will begin improving the Deal Assistant experience by executing additional features on the roadmap.
Note: If our A/B test results are negative, we will investigate potential causes, iterate, and consider re-running the experiment.
What could Deal Assistant roll out down the road?
FUTURE ITERATIONS
Custom Rejection Response feature to notify buyer they are not eligible based off the question they asked
Notify buyers how many other buyers the seller is speaking to you about the product
Give sellers the option to delay the auto-response message to buyers
Allow sellers to provide their active messaging hours
Use AI to automate responses to buyer questions that the seller did not include in their preset answers
Give sellers access to Deal Assistant by providing prepared responses if the buyer confirms the product is still available
If a buyer inquires about a product that is sold, an AI-generated message can mention similar products available
Final Thoughts
Final Thoughts
SUMMARY
OfferUp prides itself on providing a simplistic experience for buyers and sellers alike. In order to follow through on this commitment, the company must provide both buyers and sellers with as seamless of an experience as possible.
While going through this case study, I recognized the vital importance of automating features wherever automation is possible. As our world progresses technologically, especially with the implementation of artificial intelligence, it would be foolish to not implement such tools.
I hope you enjoyed this case study!