Passbase is a B2B SaaS identity verification platform, enabling companies to authenticate their users through a combination of an ID-scan and a selfie. Today we have around 30 employees and hundreds of customers operating across sectors ranging from financial services to digital health.
In Part 1 of this series, I covered some of the backstory of Passbase and how we think about the Customer Success division in the context of our business. This time I focus on the guiding principles of our CX team, analyse the importance of prioritisation and dive into the concept of a “health score”.
Establishing guiding principles
2020 was a real rollercoaster for us. We were lucky to grow significantly, but this also meant creating processes on the go and often making difficult compromises between speed and quality. I wasn’t quite sure how to summarise the feeling before seeing the perfect visualisation tweeted by one of our investors Hadley Harris:
Given the limited resources to accompany a continuously growing customer base, we are in a constant time allocation battle between reactive and proactive activities. Thankfully, over the past few weeks I’ve had some time to recap and think about the experience we want our clients to have with us.
I wrote down some principles to act as a guiding light for any activities of the Customer Experience Team:
- Tell it how it is - transparency is the only way to establish trust and for a long-term partnership to work, there is nothing more important than trust. This is especially true when you fuck up.
- Understand in depth - make sure you know exactly how the customer is using your product, how often and for what purpose.
- Act as a thought partner - even though our clients’ compliance and product teams usually define their workflow, we are in a unique position to advise on best practices and help them succeed.
- Do the right thing, even when it’s difficult - a value, that is also a guiding principle for our internal operations. Work closely with the company to identify and solve their root problem in a sensible way, rather than just the easy way.
Out of these 4, the second one is the foundation - knowledge. This means having easy access to complete information about our customer base and establishing a repeatable process of gathering data on any new onboarders. Once this layer is in place and scales smoothly, you’ll be in a good position to tackle the rest.
The problem for us was not necessarily the lack of information, but rather the multitude of sources we use for capturing specific pieces of data. For instance, our sales team uses Hubspot for tracking potential leads and information about their intended usage. That said, we found Hubspot not to be a great option for post-sales activities, due to a clunky interface and poor visualisation options for usage based criteria. Given that our payment information and product usage data come from a second and third source, we just needed a more flexible platform where to pull all of the information together.
I took a look at a couple of Customer Success specific tools, but they required quite a hefty investment upfront and for our initial short-term monitoring purpose, too much energy to set up. In general, there seems to be some room on the market for an easy access customer success tool, where you could create workflows with drag and drop integrations and pay a simple monthly subscription based on the number of users tied to the account.
Hey, maybe the next startup?
In the interim however, we decided to build a scrappy solution in Airtable, pulling all of the data together into a simple spreadsheet. We still update individual sources, but build visualisations and get answers to customer related questions (e.g. Which integrations are they using? How does X look across our entire customer set?) there.
As the transition is still a work in progress, it’s too early to tell, whether it’s a viable solution or not. That said, one of our Q1 goals for the CX team is to be able to answer all high-level customer related questions with a click of a button, so I’m sure I’ll have more thoughts on this soon.
We have the information, now what do we do with it?
There’s a famous article by a Point9 partner Christoph Janz about the five ways to build a $100 million business. To summarise, he divides target customer profiles to 5 categories:
- Flies - representing 10 million active consumers who you monetise at $10+ per year
- Mice - representing 1 million consumers paying you $100+ per year
- Rabbit - representing 100,000 small businesses paying you $1k+ per year
- Deer - representing 10,000 medium sized businesses paying you $10k+ per year
- Elephants - representing 1,000 enterprise customers paying you $100k+ per year each
In the case of Passbase, we have a significant self-service component, which often leads to customers in the rabbit and deer sections, whereas our sales team actively hunts for elephants and beyond (see Janz’s extension to the original article here).
From a CX perspective, we need to excel for all three, but as much as we would like to give everyone white glove treatment, it is not pragmatic nor plausible to do so. Moreover, clients from different categories carry different expectations. For instance, organising a monthly call to go over verification statistics might make sense for the enterprise customer, but does not for the 10 person team running a hundred verifications a month.
Hence the need to segment your customer base.
In the categorisation process, we look at three values:
- Expected volume - how many monthly verifications do they run?
- Advocacy - how much do they contribute to our customer development?
- Brand Value - how many people are familiar with the brand?
Each of these values is granted a score on a 4 point system and a segment is assigned through a formula. While the first and third value are relatively self-explanatory, you may be wondering about advocacy. One of the key propositions of our product is its ease of use and accessibility, which is perhaps why we’ve found referrals from developers and product managers using the platform to be an extremely useful customer acquisition channel.
While segmenting is oriented around revenue driving goals (e.g. higher volumes = higher revenues), we cross reference it with another criteria called the Health Score.
A customer health score represents the likelihood of a particular outcome (e.g. renewal or churn). The way to calculate a customer health score is completely individual and up to the company to define. Usually it involves some combination of a usage metric, feedback, number of support cases and engagement. In our case, we want to be sure that we are living up to our customers’ expectations, which is why we keep close account of their predictions along with actual usage. Our Health Score encompasses the following:
- Customer success manager’s sentiment - on a scale of 1–10, how happy do you think the customer is?
- Volume % - how does the customer’s volume compare with their expectation?
- MoM growth - is the customer’s volume growing and by how much?
- #of bugs escalated to engineering - how many product concerns required the involvement of our engineering team?
- NPS - how likely is the customer to recommend the service?
Again we evaluate these on a 4 point scale, but this time use colour coding (Green, Yellow, Orange, Red) to be able to quickly identify which customers are at risk of churn and where should we turn rapid attention to.
By cross-referencing these two values (segment and health score) we can start creating a resource allocation chart for our team, building sets of activities for various scenarios and establishing a thoughtful and balanced customer success department.
- Establish guiding principles for your CX department by asking yourself: What is the impression I want our customers to have about us?
- Make sure you have a process in place for gathering and storing key information about the client, so you can make informed suggestions
- Analyse your target customer profile and the expectations they have of your service: white glove treatment vs. self-service
- Categorise your customer base and build a health score that accounts for both subjective and objective criteria