In August, we audited 250 discussion threads across 25 customer support communities to find out how customers were using support forums, if they were providing the promised benefit, and how we might improve our software. This study did not consider the numerous benefits of a customer community to marketing or other company departments. The communities were both B2B and B2C and included leading companies in the following verticals: high tech, gaming, consumer durables, retail, and Telcos/ISPs. [For the full infographic, please scroll down]
Here are the results:
What is the nature of the questions being asked?
|Asking for technical support||55%|
|Asking for advice or best practices||24%|
|A complaint or suggestion||10%|
|A social or off topic post.||11%|
No surprises here. These findings were pretty consistent across all types of companies. The only exceptions was a higher percentage (24%) of off topic discussions in communities supporting consumer products where a higher degree of enthusiasm makes things more social.
There is some disagreement over allowing off topic conversations in support forums. There are some pros (ex. customer bonding) and cons (ex. noise, forming of cliques) but there is no recommended practice that applies to all communities.
Was the question itself answered?
|Was the question answered?|
29% of questions left unanswered is too high. Community managers and support staff should incorporate identification and escalation of unanswered questions into their daily routines. Vanilla’s Q&A plugin has several features to address this but, evidently, they can be improved.
A note on this question: many unanswered questions drew comments from customers or staff but no concrete answer was provided.
Who answered the question? A staff member or another customer?
|Who answered the question|
This was a pleasant surprise. The fact that over 80% of questions were answered by other customers shows that a support community can deflect a significant number of tickets from support agents. It should be noted that while these numbers were fairly consistent across all verticals, they varied from community to community. Some communities had virtually no staff presence and so questions were, eventually, answered by other customers. Other communities had very keen staff members that would jump on everything and answer questions as quickly as they could.
Community managers should think about setting a target for community contributed answers and develop practices that will help achieve that target. This could include things like letting new questions sit for a while to give another customer the opportunity to answer even if it goes contrary to our instinct to provide the fastest possible turnaround. A more sophisticated approach might also consider the nature of the question or the OP’s (original poster) profile.
How long did it take to get a response and an answer?
|Time, in hours|
|Average time to first response||31 hours|
|Median time to first response||2.5 hours|
|Average time to answer||37 hours|
|Median time to answer||6.7 hours|
While more than half of all questions were answered within a reasonable time, the averages are very high (more than 24 hours). If we remove the top ten outliers from our data, the average time to first response drops to 18 hours, more reasonable but still longer than an 8 hour business day.
Questions posted to community forums are usually more complex than those that can be addressed by consulting a knowledge base or by doing some quick online research and therefore customers are more willing to wait for an answer. That said, community managers should be escalating questions that sit unanswered for too long. Vanilla and other leading forum providers can connect to CRM and ticketing software to enable easy escalations.
What is the level of community participation?
|Average # of people commenting on discussion||3|
|Average # of comments per discussion||5|
Customer engagement can be measured by looking at participation. Here we see that support forum discussions draw in very few comments and participants compared to enthusiast and lifestyle type communities. Support communities are more transactional in nature; once a question is answered, it’s answered. We did not measure participation through Reactions (a Vanilla feature that lets users tag a post with a variety of sentiments such as ‘Like’ or ‘Insightful’) but we did observe it to be very common as it is an easy way for others to participate in a discussion thread.
How many times is a discussion viewed?
|Views (within first 60 days)|
|Average # of discussion views||87|
Customers are expressing more and more a preference for self-service support. The community is not just a way to enable conversations, it becomes a right and dynamic online knowledge base over time. The more views each discussion gets, the more second order value it is creating. We did not look at where the views were coming from for the specific list of discussions in this study but we know that the majority of traffic comes from search engines. Making sure your community is SEO optimized and also having a good internal search engine is key.
It should also be noted that the number of views will depend on the size of your customer base and whether the forum is behind a customer only login or open to the public. Our sample included companies with tens of thousands of customers and some enterprise vendors that had customers in the hundreds or low thousands.
Was the discussion title descriptive?
|Descriptive discussion title|
The more descriptive a discussion title, the more likely it is to be answered by another customer, found through search, or viewed by someone perusing the community. The stats show that customers are very good about discussion titles. If this is an issue in your community, you can make a note of it in a dismissable message or your guidelines post or even manually edit bad discussion titles.
Did the Original Poster (OP) acknowledge the answer?
In Vanilla, an answer can be acknowledged in many ways: a written comment, clicking on a Reaction, or marking a comment as having answered the question. An acknowledgement or thank you reinforces member participation. To increase this metric, community managers can acknowledge responses on behalf of the OP and make use of badging and gamification features to reward them.
What was the tone of the discussion?
Article comments, social media posts and enthusiast forums can create a false impression that any open platform is going to become a cesspool of negativity and harassment. We found that even in video gaming communities, this is not at all the case. Our theory is that when customers are posting amongst peers (customers of same product or service) they are less likely to lash out.
This sample of support communities show that the promised benefits are in fact real. Communities also offer marketing benefits such as creating loyalty, trust and awareness amongst customers and prospects. Areas where community managers and Vanilla can improve are in finding ways to reduce response time and increase the acknowledgement rate to reinforce peer participate.