In this article
- Background
- Why SysCloud built a customer success framework on HubSpot Service Hub
We Unlocked HubSpot Service Hub to Catalyze Customer Success and Reduce Churn
20 Mar 2023
|5 min read
|Annu Sabu
Blog Articles
Here is how we transformed HubSpot Service Hub into a customer success platform without using any third-party customer success solutions. Read this article to download our customer success model and HubSpot workflows so you can upgrade HubSpot Service Hub to a full-featured customer success platform.
Background
There are two fundamental reasons why SaaS companies should go beyond customer service and think in terms of customer success.
- Boost product adoption by demonstrating value with minimal friction
- Maximize product engagement to give customers a reason to stay
So how does a growing SaaS company graduate from customer support to customer success?
Using customer service tools like Jira, Zendesk, or even HubSpot Service Hub may not always cut the ice.
At SysCloud, we too felt a need to graduate from a reactive customer service approach to a proactive customer success model.
In order to make the transition, the first approach that we zeroed in was as below.
Preliminary approach to customer success implementation at SysCloud
But this approach had its drawbacks:
- Even though Freshdesk had some interesting features like screen share in chat and offers KB articles integrated with chat, we needed deeper integration with our database and wanted to run automated workflows from the tickets.
- Google Analytics can only track individual user actions and behaviors like clicks, page visits, and login frequency and there is no way to link user behavior to accounts. Even the latest GA4 version falls short on this front.
- When we evaluated third-party customer success apps, we felt that implementation would require complex data integration with our application.
We were already relying on HubSpot Marketing Hub that was integrated with the SysCloud cloud backup product via the HubSpot API.
Running parallel integrations with customer support and customer success platforms was a recipe for disaster for growing SaaS company like ours.
There are two main challenges that can complicate the implementation of a customer success solution.
1. One too many data flows and connections: Integrating a new customer success solution would mean data is constantly moving from our application to HubSpot to update the CRM. In addition, we need to make a similar integration between our application and the customer success platform. Even a single instance of an API error will have a cascading effect on the quality of data and reports. Troubleshooting mismatched data can also become cumbersome.
2. Generating cross-functional reports is difficult (figure 2): To create a customized report in HubSpot, data must be mapped from the customer success platform with components in HubSpot which in turn may be populated from our application. This seemed like a tedious process and any mistake in mapping could lead to incorrect reports.
We managed to substantially reduce the integration and reporting complexities by subscribing to HubSpot’s Service Hub as you can see in the diagram below.
So, we decided there was a better way to move forward.
Why SysCloud built a customer success framework on HubSpot Service Hub
We were already using HubSpot marketing cloud, and it dawned upon us that the easiest path to creating a customer success platform was to build one on HubSpot itself!
This might sound like a far-fetched idea, but we had objective reasons.
HubSpot vs. customer success platforms
HubSpot positions their Service Hub as a customer success platform. However, it may not have all the bells and whistles offered by third-party customer success platforms as shown in the table below.
The comparison table below is just the tip of the iceberg.
Though HubSpot lacked customer success features it had more to offer in terms of CRM management and Marketing tools, as you can see in the table above.
How we built a customer success framework on HubSpot
The in-house customer success framework built on HubSpot required the following steps:
1) Conceptualize customer health score model
2) Create custom properties on HubSpot
3) Populate HubSpot properties
4) Build workflows to notify customer success, sales teams, and product managers
5) Create reports and dashboards
What is a health score?
Every customer is assigned a health score to denote product adoption and satisfaction. The health score reveals whether the customer is happy with the product and is also a leading indicator for upsell, cross-sell, and churn. Health scores also help customer success teams to prioritize their time based on objective data.
How to create a health score model on Service Hub
Step 1: Identify the health score components
The first task is to frame the formula to compute the health score.
While it may not be possible to come up with a universal health score model, here is a simple framework that most SaaS companies can use to figure out the formula for the health score.
List down every customer touchpoint right from account creation, the entire free trial journey, and the process to upgrade the subscription. Include expansion and cross-sell journeys as well.
The weightage given to each section depends on the impact it has on customer experience.
| Factors | Sub factors | Score | Rules | Score of data |
|---|---|---|---|---|
| Onboarding (15/5) | Onboarding completed 1.1 Without issues | 15 | After 3 months these 15 marks will be allocated to the app health status equally for all the accounts. This is because onboarding is a one-time event. So, the positive or the negative experience is discounted after 3 months. | Application and HubSpot |
| 1.2 With issues | 5 | |||
| Business value (15/15) | 2.1 (Your product USP) Example: Number of restores/exports in the last 3 months | 5 | Consider this component for calculation only if there is a login activity by a super admin or user (depending on if your product is B2B or B2C) in the last 3 months/90days. Licenses utilization should be more than 70% of the purchased licenses. | Application |
| 2.2 (Your product USP) Example: If the product home page was viewed in the last 3 months. | 5 | |||
| 2.3 Licenses purchased vs Licenses utilized | 5 | |||
| Tickets (10/10) | 3.1 Tickets not submitted | 10 | Consider this component for calculation only if there a login activity by a super admin or user in last 3 months/90days. Tickets submitted in last 1 month will be taken into consideration for calculating the score. | Hubspot |
| 3.2 Tickets submitted & resolved as per SLA | 7.5 | |||
| 3.3 Tickets submitted & not resolved as per SLA | 5 | |||
| App health status (42.5/42.5) | 5.1 (feature) Example: Backup completed within 24hours | 15 | The application loading time is scored only if it is loaded in 4 seconds. The temporary errors are scored if they are resolved within 24hrs. | Application |
| 5.2 (feature) | 12.5 | |||
| 5.3 (feature) | 5 | |||
| 5.4 Application loading time | 5 | |||
| 5.5 Temporary errors | 5 | |||
| Relationship (15) | 6.1 Quality of touchpoint | 5 | Email or phone conversations with customers, 1. In the last 2 weeks is marked "Very Good" - 5 2. The last touch is before 2 weeks and earlier than 4 weeks is "Good" - 2.5 3. Anything older than a month is "Bad" - 0 4. No conversation in 1 quarter is "Non-Existent" - 0 | CSS executives manually feed data in HubSpot, or using feedback forms |
| 6.2 Perceived satisfaction | 5 | Customer satisfaction levels, - If satisfied with the solution - 5 - If unsure of level of satisfaction - 2.5 - If needs immediate attention - 0 | ||
| 6.3 Blockage | 5 | Blockage to reach the customer, 1. Hard to reach - 0 2. Unmet feature request - 0 3. Point of contact left - 0 4. No secondary contact - 0 5. NIL - 5 | ||
| Unexpected changes (10) | 7.1 No unexpected changes | 10 | No specific rule | CSS executives manually feed data in HubSpot |
| 7.2 Company Restructure | 0 | |||
| 7.3 Request to cancel | 0 | |||
| 7.4 Acquired | 0 | |||
| 7.5 Budget constraints | 0 |
Step 2: Create custom properties
Once you have defined the health scope formula, the next task is to identify the HubSpot properties that would give you this data.
Some data might be readily available as out-of-the-box HubSpot properties such as name of the user, domain name, country, and region.
For the rest, you may have to import them from your application to custom HubSpot properties using HubSpot API calls.
For example:
- To know how many SysCloud licenses were used by a customer, we created a custom property named ‘Backup licenses used’.
- To know if there were any tickets created in the last 30 days by a customer, we created the property ‘Tickets raised last month’.
Pro tipAll our health-score related properties are defined for the object ‘Companies' and the scores are calculated for each company and not contacts. We did this to avoid multiple scoring for the same customer when more than one employee from the same organization is enrolled in our HubSpot database.
Step 3: Update the properties on an ongoing basis
This is the most time-consuming and maybe the most complicated part of this process. There are three ways to add data to the health score properties:
1. API calls
Properties that needed to be updated with data from the SysCloud application were fetched using HubSpot API calls. Refer to HubSpot’s API documentation to know more about integrating your application with HubSpot and fetching data through API calls.
2. HubSpot workflows
A few properties had to be updated using workflows. Here is an example of such a property.
We created a custom company property named ‘Tickets raised last month’ to know if any support tickets were created in the last 30 days. The data under this property would be a simple yes or no drop-down. Even though you can see all the tickets raised by a contact or a company, HubSpot does not automatically tell you if they have or have not raised a ticket in a certain period.
So here is what we did.
We created a ticket-based workflow, to check if a ticket created in the last 1 month was resolved as per our SLA or not. If a ticket was not resolved as per our SLA, the associated property will be marked closed ‘Not as per SLA’.
Here is an image of the workflow.
Similarly, we created other workflows to check if the user has logged in to our application in the last 1 month or not.
We also created tasks based on these workflows to actively follow up with leads.
If a user had not logged into the application for over a month, an email was sent to notify the respective customer success service agent. simultaneously, a task was also created and assigned to the customer success service agent.
3. Manual entry
A few properties must be manually updated by the respective customer success service executive (customer success service executive). for example, in the ‘relationship component’ of the health score model, the customer success service executive must manually update the properties under perceived satisfaction, quality of touch point, and blockage.
We then created workflows based on these manual updates to check how satisfied the customer was with our product. If this property was not marked in HubSpot, a reminder email was sent to the respective customer success service executive through workflows.
Step 4: Computing the real-time customer health score
Once we had all the properties populated, we gave each component a score based on the scoring chart we created.
The health score algorithm was created in our application database.
Contact us if you need the full JavaScript code that we used to calculate the health scores.
The final scores were then synced to the HubSpot property called ‘Backup health score’.
Once the scores were assigned, we categorized the scores into high, medium, and low. We created another company property called ‘Backup health score priority’ and ran a workflow to automatically update these properties.
In this workflow, if the score is between:
- 0-50 the property is marked as low.
- 51-74 the property is marked as medium.
- 75-100 the property is marked as high.
As we wanted a fluid scoring system, the properties are updated daily in real-time using a different set of workflows.
Step 5: Create reports and dashboards
Finally, we created in-depth reports and a dashboard to visualize our health scores.
Here is an image of the health score dashboard we created.
The dashboard would show you how many customers are considered to have a good score and how many are on the verge of churning.
Here are the data components we used to create this report.
Key takeaways from our experience
HubSpot should be seen as a “platform” and not a “product”. There is a fundamental difference between the two perspectives. While we initially subscribed to HubSpot for its value as an inbound marketing product, we realized that we could gain better ROI if we treated HubSpot as a platform.
Implementing custom solutions on HubSpot requires technical chops. So don’t try to follow our model if you do not have engineers who can work with HubSpot APIs and configure advanced workflows or reports.
Remember to budget time and effort to maintain all integrations with HubSpot. This is not a “build it and forget it” solution. It requires constant upkeep and periodic troubleshooting to fix API errors.
While it is easy to run workflows and generate automated tasks for your customer success teams, it is important to take your customer success team into confidence before rolling out a customer success solution. If your customer success team starts ignoring the automated tasks and notifications, it defeats the purpose of this solution.
In this article
- Background
- Why SysCloud built a customer success framework on HubSpot Service Hub