At InVideo, when we launched on April 10th, 2019, our product was not launch-ready. In fact, we were going through a pivot. Long story short, a broken product during a one-time campaign that will send us thousands of users within a week was scary, very scary. Sailing through it successfully was crucial for the survival and CUSTOMER SUPPORT came as our lifeline. Since then, we have strived to provide exceptional customer support.
Our customer support replies within 2 mins 24*7 via chat for free as well as paid users.
Here is a screenshot of all folks from our support team with their first response time and a few other metrics.
This kind of customer support is unheard of, especially in a prosumer SAAS companies where ticket sizes are ~$10/mo for paid users and a conversion rate of only 3%. However, ‘Why did we start 24*7 customer support’ is for another day. Let’s focus on how we use customer support as a product management weapon.
Some context first:
Today, we handle ~500 customer chats/day. That is like talking to 15000 customers every single month, 1:1. Let’s see what we made out of it.
We bifurcated the support chats into Product Help, Feature Request, and Bugs. And then we went deep — Here is a screenshot of how the bifurcation (tag) of every chat looks like. Here is the read-only file, in case you cannot see the screenshot.
We are also very disciplined to add new tags during a new feature release. That helps us keep the tags/taxonomy sane. Let’s see how deep we went with the tags — Example:
At InVideo, we call this as a left panel:
It has several components like music, text, effects, etc. as you can see here
Now, let’s look at how deep our tagging mechanism goes.
and then level 2:
Basically, we are going to the root of the problem. The problem cannot just be ‘Image’ related, it needs to be further analyzed. And all this data is passed on to the product team.
Tagging. Ensuring that every chat is tagged right by our customer support team. Now come on, we are all humans and we all make mistakes — But hey, they are mostly accurate to 85% of the time, and that helps us with the best indicative feedback.
The weekly report from our head of customer support:
We mainly focus on the percentage bifurcation of our macro-categories like Product Help, Bugs, and Feature requests. Then we go deeper. Here are some of the examples:
(This image is from a product we just started experimenting with — Aavaz, I am excited and their alpha version looks promising.)
Basically, we find out if there are specific spikes in a particular category very easily using this mechanism.
We further go deeper to identify specific issues — Here are a few examples:
The ones related to text —
Or my projects page —
Or feature request on a left panel
Or feature request on the canvas
This truly helps us understand what our users want, where are they suffering, and helps the product team to evaluate priorities through talking to 15000 customers/month.
Our support presentations are 78 slides long, here is the link
A lot of our decisions are driven from the feedback we get via support be it a response to the new release to solving a bug to improving the UX.
We look at metrics like the total number of chats related to bugs, which feature gets the maximum number of product help queries, what are the most used support articles etc: Here is a screenshot of our most viewed articles in the last 7 days. It means that these features are difficult to understand and PM’s need to solve them.
Good customer support is a moat.
If we find an issue with a customer, we do not shy away from going all out. Here is one of the screenshots of that.
Friday at ~1 AM —
When I look back to see how obsessed the team is, I feel privileged.
At InVideo, we believe we are changing how bottoms up SAAS companies are built for the world. We at InVideo strive to provide a consumer-grade product, enterprise-grade sales, service and support, and prosumer pricing. All this while maintaining healthy gross margins.
I am deeply grateful to everyone who took support chats through the nights in April-July 2019 when we did not have the support team, worked in erratic shifts. I also am very grateful to Hrishikesh, Swapnil, Hemangi, Pratik, Akash, and Brandon who have been the backbone of our customer support. It was just not possible without any of you folks. Thank you.