How we scaled customer delight and got to 18,000 users in 2 years via word-of-mouth
Goalwise is a personal wealth management startup based in India — now more than 2 years old and with over 18,000 users (mainly acquired via word-of-mouth). Our growth strategy is pretty much based on one mantra — the customer experience should be great!
Basically that means that everything that the user touches should delight them or at the very least not annoy them given the industry we are in. Typically, if a user is delighted about something they will tell maybe a maximum of 5 friends but if they are annoyed with something then they are likely to tell at least 5 friends.
The Problem
We were not even a year old and we had a problem. Not all problems are bad. Some, as they say, are good problems to have. We had a ‘good’ problem — the number of users were growing faster than we had estimated. This resulted in a small team (less than 15 in total — including tech, advisory, customer support, operations etc.) trying to service more than 4000 users in a technologically backward industry with lot of offline processes. And of course we weren’t really prepared for it — no one is.
This good problem resulted in 3 bad problems:
- Things started to ‘break’ — Things start falling through the cracks because we weren’t really used to servicing so many people. For example delay in verifying documents and activating accounts, not being able to answer chats immediately because you are also on a call with another user and other similar things. These were things that we could get past by apologising to our users but these are exactly the things that prevent us from truly delighting our users. And that was a big problem because that was our ‘growth strategy’.
- We didn’t know which problems to prioritise — There were many things that needed attention. Most of the problems were being fixed on the spot as and when they came up. Some of them were recurring issues but no one was tracking them so that later they could be fixed once for all. This was mainly because there was no time to take a step back, take stock of the big picture and think of scalable solutions.
- Loss of accountability — We couldn’t fix accountability to departments/people because we did not have a process to document all the things that were going wrong in the first place. Heck, we did not even have proper documentation of processes to be followed especially when tasks involved inter-departmental work. For example, if a user has not submitted a required document because of which his account cannot be activated, who should remember to follow up — customer care or operations? This just led to more frustration amongst everyone with no actionable points.
What did we do?
Although at the time, the issues were not as clear in my head as they are now in hindsight, we knew that small things were going wrong, we were getting sadder with every apology and we had to start improving somewhere. So I came up with a simple sheet called the User Happiness Sheet (UHS).
It had 2 kinds of entries made as and when they happened -
Sorry entry — When we said sorry to a user for a mistake we had committed or a bug in the product. For example, if a user could not upload his document properly would count as a sorry. If the error message was confusing, that would be counted as a sorry as well. If we had told a user that we will get back to them by end of day and did not, then that would be counted as a sorry. But if we did not have a feature because we did not think it was important and a user complains of it, it wasn’t counted as a sorry (but was passed on as feedback for product).
Thank You entry — When a user complimented us enough for us to say Thank you to them. For example, a user sent us a lovely printed letter saying how good he feels about investing with us. That counted as 1 thank you. But if the user just said something like ‘Thanks for the help’ it wasn’t counted since we wouldn’t be replying with a thank you to this. So it had to be a compliment to which we say ‘Thank you for your appreciation’ or something like that. All 5-star reviews on Facebook and Play store were also a part of it.
For each entry, we noted the date, the userid, the exact words of the user leading to this entry and which department would this be attributed to.
For example, delay in account activation would be attributed to the ops team, incorrect info provided on chat/email would be attributed to the care team, bugs would be attributed to the tech team, UX issues would be attributed to the product team etc.
The same would be done for thank you’s as well. If the user was appreciative of more than one type of thing (eg product + customer support), then it would be attributed to both. And some entries were attributed to Team Goalwise. For e.g. a 5 star review on Facebook praising us for our overall experience.
Since all the communication went through the customer care team, they were responsible for making the UHS entries for themselves as well as other teams.
And every week we calculated the Thank you:Sorry ratio aka our User Happiness Score. When multiple departments were responsible for a sorry or thank you and then 1 point would get split amongst them. So this gave us a User Happiness Score at the overall level and each department level. We also had columns which explained exactly what the compliment was about or the mistake was.
The sheet was accessible to everyone in our start-up and not just the department heads. Everyone could see what was going on. We made it a discussion point in our department and all-hands meetings.
The first week we did it, we realised our hunch was right. We were saying way more sorry’s than thank you’s. The Thank you: Sorry Ratio was 0.55 i.e. we had almost 2 sorry’s for every thank you. It was disappointing especially since we really wanted to build our brand on user delight.
It’s important to note here that all our mistakes would go into this sheet as a Sorry even if the user was not really pissed off because of it. But to earn thank you entries, we had to do delight a user enough that he gives us a compliment. So it was difficult to get a high ratio especially in the phase we were in almost by design, but anything worthwhile is never easy.
And we could see that the issues were not things that could be fixed overnight. We needed to automate a lot of processes internally to decrease manual errors, improve our QA before we shipped features to reduce bugs etc. All of this required time and investment (and leadership).
But we kept recording and measuring and discussing till it became common vocabulary.
And then something happened.
We started improving just because we were measuring transparently.
Manual errors started going down as people became more aware of how their work is going to be evaluated — by the UHS yardstick. Then they started pushing for documentation and setting up processes on their own rather than it being forced from top down.
We set a target for ourselves — to get the Thank You-Sorry ratio to 1 in the next quarter. This really brought all departments together in a mission mode and everyone started looking for ways to eliminate sources of sorrys and to increase customer delight. In a sense, it became our north star metric across the organisation.
We analysed everything that was going wrong and of things that were earning us thank yous.
Usually they fell in one of the categories below and could be attributed to a lack in our processes:
Sorry’s:
Manual errors
- Lack of automation in internal operations processes
- Lack of documentation of processes
- Lack of processes themselves
Bugs
- Lack of testing
- Lack of proper requirements from product before the tech team started development
Incorrect communication, missed follow-ups
- Lack of knowledge
- Lack of accountability
- Lack of tracking communication
Thanks You’s:
Knowledgeable and friendly customer support
- Focus on training of customer support
Honest advice
- Focus on company values
Great Product
- Focus on design and usability
Every week all unit heads would go through the sheet, look at all the new entries, trace it to the underlying cause and come up with a solution. This list of solutions would then be discussed and prioritised and put on our roadmap. This went a long way in fixing accountability. And of course things that were breaking more frequently or/and were easy to fix were prioritised first.
Instead of just adding new features, tech spent a good part of that quarter in automating internal processes. We built two internal products — one for care and one for ops — implementing workflows unique to our industry.
Not all problems and solutions were technological though. Most of them involved setting up well defined processes and giving people clear ownership (and hence accountability). Sometimes it also meant saying no to a user if it meant making an exception to a tricky process (and there were quite a few of those initially).
And all departments worked hard towards it and we reached the target of UHS of 1 in just the next 2 months!
Even if one department hadn’t put all it’s weight behind it we wouldn’t have been able to achieve it.
Over the past year, our User Happiness Score has gone from 0.55 to 1.72 even as our user base has grown multi-fold. We have been able to keep the number of sorry’s under check and increased our thank-you’s even though we keep experimenting with our product and processes with a small team.
This is reflected in the external world as well — currently we are one of the highest rated Mutual Fund investing apps on Google Play (4.6) and Facebook (5.0).
Our User Happiness Score has been instrumental in helping us come this far.
We could have used NPS to measure how we were doing (and we did use it) but UHS gave us a much more granular and actionable look into what was going wrong and how to fix it and we were able to rally our entire organisation around it.
