How we boosted Ranking Raccoon's response rate to over 70%

Ranking Raccoon is the most recent product we built and launched at UX studio. It's a spam-free platform for ethical marketers and SEO professionals to build high-quality backlinks. Currently, 70% of link requests sent on Ranking Raccoon receive a response within a week, and 35% result in a published link, significantly outperforming methods like cold emailing, which typically sees a success rate below 10%.

But these impressive stats haven't always been like that. Shortly after launching the product in 2023, half of link requests sent on Ranking Raccoon would go unanswered, a problem our team hadn't anticipated. 

Community engagement is a crucial factor not only for users to be successful in link building but also for the product to thrive. Therefore, we needed to figure out why the response rate was so low and take action to address the issue. 

In this article, we'll walk you through the entire process and how we successfully raised the response rate to over 70%, a figure we're extremely proud of.

The challenge

Meeting OKRs as a freshly launched product

The response rate is an essential metric in link-building outreach. A higher response rate not only suggests that the approach is effective but also increases the likelihood of securing links.

On Ranking Raccoon, users can browse other verified sites that are related to theirs and contact site admins directly via chat. Given the directness of this approach, we didn't anticipate that the response rate could be low, especially since messages on Ranking Raccoon are focused and not buried among thousands of automated emails in a flooded inbox.

However, we overlooked one detail: our target audience was not yet used to being on Ranking Raccoon every day. It was a new product, and marketers already have to multitask and juggle many different tools and responsibilities on a daily basis.

Locating the issue: data-collection and user interviews

But how did we know that the response rate was low? We have an established practice of inviting users for a chat every week and discussing their experience using the platform. This was our very first sign: users told us that they tried the tool and really liked the concept, but in practice, they were not receiving responses to most of the messages they sent. So, if they were already not visiting Ranking Raccoon every day, they now had reasons not to return at all.

To make our assessment more objective, we sat down and examined the data: we track how many requests are sent in a week and how many receive a response. The response rate was in a declining trend, and we hit our worst week at 35% response rate around the time when we realised we had a problem.

In any given week, there can be external factors influencing low engagement, such as long holidays and events that many professionals in the field attend. However, the fact that there was a downward trend in response rates over time raised a red flag for us, making this the main topic of our weekly product meetings.

Setting problem-focused goals and KPIs

We were so determined to solve what we called the “ghosting problem” that we set it as our quarterly objective, prioritizing it above anything else due to its significant impact on the user journey and the product’s success.

The goal that we set for ourselves was to ensure that at least 75% of new requests received a reply.

Understanding the problem in-depth

Although we're now proud to say that we have achieved a 75% response rate, it was a long journey filled with moments when the problem was far from clear. 

We had multiple assumptions regarding why users were ghosting one another, but we all agreed that we needed to identify the actual reasons before attempting to solve the problem with experimental solutions.

Initial assumptions

Here are some of the assumptions our team came up with to explain the low response rates:

Notification issues

Maybe users were not properly or sufficiently notified about new messages.

Overlooking messages 

Users read messages but forgot to reply. If this was the case, then our product interface didn't do a good enough job reminding them of unanswered messages.

Communication style

Our users are used to email communications and may find it difficult to decline requests so directly in a chat format. In email, not responding is often seen as acceptable, so perhaps we were not facilitating friendly resolutions for open requests.

Lack of guidance

Maybe we didn't sufficiently educate users about our new approach to link building, failing to encourage users to engage and build relationships that could lead to links in the future.

Inactive users

We hypothesised that new users who presented with all sites on Ranking Raccoon, some belonging to users who were not active for a considerable amount of time, leading new users to message people who were not likely to respond from the start. We did not provide indications of who is more likely to reply or warnings about those who are less likely to engage, which may have led to low engagement rates.

Overwhelmed marketers and SEOs

Our target audience often has to juggle too many tasks at once, so maybe they were too busy to respond timely.

Data analysis

We checked our assumptions against the usage data to get a clear picture. When analyzing the number of unanswered messages that were read versus those that were not, we realized that many of them were actually read by the recipients. Also, we noticed that a considerable number of users who were “ghosting” hadn't been active for a significant period of time (over 2 weeks).

We also came up with several questions to ask users in our feedback chats. While the difficulty of saying no emerged as a recurring theme, we ultimately realized that there was no single cause for the problem. Each of our assumptions could be true in different situations, so we began brainstorming solutions for each specific issue.

Solutions

It's important to note that while many of our solutions were implemented successfully, our team remained a little skeptical about how much we could influence human behavior through product features. 

However, we understood that we could only learn by trying. Therefore, we compiled a list of solutions to test. We categorized these solutions into two groups: one aimed at better directing users who were not receiving replies, and the other focused on encouraging responses from those who were unresponsive.

For users not receiving answers

Activity indicator

We decided to guide users towards success in getting replies by indicating, in advance, how recently each other user has been active on Ranking Raccoon. By messaging individuals who had been active that week, users would be more likely to receive a response.

An important consideration was to avoid overly exposing user activity, so we implemented general labels to provide a clear indication of how active someone is without revealing the exact times they were online.

Sent/Seen receipts

We decided to implement receipts for every message sent by the user, allowing them to know whether their message was seen. This feature helps users understand why they may not have received a reply; it could be that the recipient hasn't read the message yet. 

This could also be effective in:

• Encouraging users to send a follow-up message in case a message was seen and not replied to.
• Informing that the recipient might need more time to catch up on their messages because they haven't read it yet.

For users not responding to messages

Focused and extended email campaign

We enhanced our email notifications, keeping in mind that we still had to win a space in users’ daily toolset and routines. To achieve this, we notify users about unread messages once a day for 3 consecutive days.

Additionally, we launched an educational campaign highlighting the importance of engaging with their messages. For up to two weeks, users were constantly encouraged to return and reply.

Indicator status and suspending inactive users (with a chance of reactivating themselves anytime)

Our user interface features a raccoon indicator that informs users about the health of their engagement in the community. It starts with the “All good” status when everything is well and users get some tips on what the raccoon is watching for. We decided to leverage this familiar aspect of the UI to track response rate.

If users hadn't replied to a message in over 3 days, the raccoon enters into an alert state, reminding them that they have unanswered messages. This alert remains active for up to two weeks, and we send them multiple email notifications in this period. 

If the user still hasn’t replied by then, the raccoon gets sad and transitions into an error state. This change is directly linked to another measure we introduced: after 2 weeks of not replying to a request, users will get their accounts suspended. What does suspension mean? Their websites are not visible to others and they can't start new chats, but they still have access to their accounts.
This measure was somewhat controversial, but we implemented it to protect new users from contacting individuals who are unlikely to respond, given their lack of replies in over two weeks. As a result, suspended users' sites are hidden from other users’ dashboards, and they will no longer receive new messages.

While this approach may seem drastic, suspended users can easily resolve the issue and restore their profiles to normal. They simply need to reply to their unanswered messages to regain access and have their sites reappear on other users’ dashboards

Highlighting unanswered requests (banner, tags, in-chat messages)

Users may have multiple conversations happening simultaneously on Ranking Raccoon, so we incorporated several UI elements to direct them to conversations with unanswered messages.

One-click answers (high adoption)

Ultimately, our goal was to encourage users to respond even when they couldn't provide a link to the site of the person requesting one. To facilitate this, we introduced a set of one-click answers that users can easily select, ranging from Yes to Maybe to No. This way, regardless of the outcome, users can still send a reply.

We also aimed to address the needs of busy users who may not have time to frequently check their messages on Ranking Raccoon. To accommodate them, we included an option for “I’ll get back to you shortly.”

This feature has been widely adopted, and we are proud to have introduced it, as it has benefited both parties involved in Ranking Raccoon chats.

What didn't work well

While we have highlighted the strategies that were successful, it’s important to acknowledge that some initiatives did not perform as well.

One of our strategies to encourage replies was to celebrate a new incoming request. However, our analysis revealed that this approach had no significant impact on whether users chose to reply, leading us to remove it from our UI.

In addition to addressing unsuccessful attempts, we also needed to iterate on the elements that were working. We realized that our frustration with the ghosting issue had caused us to adopt a tone that was overly direct in our messaging, which negatively affected the user experience. After gathering feedback from users, we revised to a more instructional and supportive approach.

What success looks like

After implementing these changes to the product, we began to notice an increase in response rate. However, we wanted to ensure that this improvement was not just a temporary spike; we needed to monitor the data over a longer period to determine if the high response rate would be sustainable.

We continued tracking weekly data, and it consistently increased. Initially skeptical about our ambitious goal of 75%, we were thrilled to achieve an impressive 86% response rate during one week following releasing our solutions. Additionally, user feedback improved significantly, and the response rate was no longer a recurring issue in conversations with them.

Now, nearly one year after addressing the situation, we're proud to report that the average response rate has stabilized around 70%. This achievement has become a key differentiator for us, especially when compared to cold emailing, where the average response rate is around 5%. As a result, it has been prominently featured on website pages and integrated into our marketing strategy.

Conclusion

The response rate case taught us that proactively addressing user engagement issues can lead to significant success. Even without having all the answers from the start, doing your best as a product team to understand the underlying causes and experimenting with potential solutions can be highly rewarding.

This experience also highlighted that solutions to pressing problems may take time to demonstrate their effectiveness and impact. It’s essential to allow new features sufficient time and exposure before drawing conclusions about their success.

We'd also like to highlight the importance of being so closely in touch with users to hear from them what might not be working on your product, especially if it comes to a new product. Observing how user reactions evolve over time as solutions are implemented is crucial.

We encourage other teams facing similar user engagement challenges to adopt a similar approach: move from assumptions to understanding the problem space, then ideate and implement solutions. It’s important to recognize that the path forward may not always be clear, particularly when issues come from multiple instances. This often requires putting together a combination of solutions rather than relying on a single fix.

Need help? Request an audit, so we can help you get to the root of your issues.

‍