Support automation: the next thing you’ll need to think about to scale your API
As more companies roll out APIs for global audiences, across different time zones, languages and target markets, building out an effective, customer centric and scalable support function has become necessary, more complex and expensive.
The Hitch API Assistant has been built to provide support for growing businesses to increase speed, quality and scalability, but also automate a key internal process: creating and updating your API knowledge base. Does this sound too tactical? Keep reading.
First, a word for support managers: why you need to segment and automate
One of the challenges around scaling support, and this goes well beyond APIs, has to do with segmentation. If you want to scale, you simply cannot serve your whole customer base with person to person support. Nor will all your customers want to consume your support resources in the same way.
It has been proven over and over again that just like we consume information in different ways, we also like to get help from companies in through different channels and resources. Some will prefer to talk to a person, others will rather communicate via email, and many will just want to look for what they need themselves.
For support managers this means three things, basically: spending some time on segmenting customers based on their value (in terms of revenue), designing a multichannel support machine that addresses these different segments’ needs, and making sure they have up to date self-service support resources.
(Make sure you’ve done another important exercise with your whole API team before you build out your support function: understand the cost of your API)
Here is where support automation comes in the picture. You’ll need to consider how to automate a lot of your support processes to offer fast support to users while scaling your internal processes and teams.
Support automation has a huge impact on some of the metrics that keep support managers awake: first response time and resolution times. If you want to scale, at one point you will need to consider automation options.
Killing two birds with one stone with the API Assistant
The main objective of the API Assistant is to fully automate 1st level support, so consumers can get answers to their questions in real-time, and quickly get them escalated to a support agent if the answer they receive is not helpful.
Our customers report that between 75% and 80% of questions are answered and require no further follow up. So this is already a 4x to 5x increase in capacity to address questions in real time.
But what if you could improve and update your API knowledge base as you resolve these escalated issues? This is the assistant’s second goal, to help you maintain your KB automatically. With almost 0 effort.
The way our support system works is the moment a question gets answered by a support agent, it ingests the answer back in the API Assistant’s “brain”. There are two important outcomes associated with this: first, the next time a consumer asks the same question, the chat bot will be able to provide the answer, and second, your KB will have more accurate content.
This last outcome is critical for many support teams, who sometimes struggle with keeping their documentation up to date, which impacts negatively on the community’s perception of the API. Streamlining this process is huge for them.
Automating support does not mean sacrificing quality
The “magic” of the API assistant is that responses from the providers’ support team are continuously ingested into the knowledge base, feeding a continuous learning loop. This responds to a paradigm shift in customer service which many successful companies have already embraced: knowledge-centered support, which puts an emphasis in content and not just support interactions.
You’ll also be able to see a qualitative improvement in your API knowledge base with the help of our new Support Analytics, which will allow you to understand which topics are more important to consumers, and whether you need to spend some time on improving your core docs. Maybe your technical writer will need to create a separate guide for a specific use case, or improve the Hello World article on your website. At the end of the day, our analytics help you get unparalleled insight into the developer experience you are providing.
The API Assistant will not eliminate human intervention in updating your documentation altogether, but it will certainly automate a big chunk of it, and give you insights to spend your resources smartly.
With this approach, your team will be continuously addressing tickets but also scaling your support function, i.e. maximizing the outcomes of their time spent addressing tickets.
Common Use Cases for the API Assistant
What’s great about the API Assistant is that it can be used by small and large teams. It’s so simple that it can be embedded on your developer portal by adding a two line JavaScript snippet, or implemented on your preferred Slack channel.
A typical use case is companies that have a single support person answering questions from the community, with a limited knowledge base, and struggle to service a growing audience. Our system allows them to have on-demand support, in real time, so that the single support person only needs to address tickets which are raised for second level support, ideally higher value customers. But thinking longterm, implementing the API Assistant also means that all time spent on answering tickets starts to build out the API Knowledge base which can be continuously developed through the support cycle.
Another common use case is established teams that need to scale. An example around this would be launching a new API platform, which will most likely come with specific challenges for support teams in terms of the volume of conversations to be handled, agent’s need for training around new functionality along with a certain expected learning curve, and the risk of low customer satisfaction and poor response times. In this case, the assistant will allow the team to automate responses and avoid being flooded with tickets, while implementing a cost effective way of providing 24/7 support during the initial launch phase.
Finally, the API Assistant also helps API teams easily support their free users and invest more resources in revenue-driving customers (and that’s why I mentioned segmentation as one of the key things you need to do when building out the support function for your API).
How to get started
It is very easy to get started with the API Assistant. All you need to do is:
- Create an API profile on Hitch
- Load your machine readable documentation
- Add some basic information to your profile like an overview and FAQ (your initial knowledge base in combination with the reference)
- Turn on the API Assistant
- Add the assistant to your developer portal (this is optional)
And then you can have your realtime API support channel up and running in no time.
We’d love to hear how companies out there are supporting their API communities. Got an interesting story to share? Give us a shout on Twitter.
If you like this article, please click on the ❤️ button below and follow this publication so you don’t miss out.
Ready to start engaging and growing your API community with Hitch? Sign up now.
