Delayed ROI signal from licensed products — Victoria Zhang // Salesforce

Victoria Zhang, VP of Data Science at Salesforce, explores customer centricity and product performance. In the realm of consumer products, our actions and experiences often have immediate and trackable effects. However, in the world of SaaS, due to the license-based business model, making the connection between daily user experiences and future revenue is a nuanced challenge for analysts. Today, Victoria discusses why data-driven SaaS product development drastically differs from consumer products.
About the speaker

Victoria Zhang

Salesforce

- Salesforce

Victoria is VP of Data Science at Salesforce

Show Notes

  • 01:01
    SaaS business models and delayed ROI
    The two most popular models are consumption-based and license-based. Theres delayed ROI in a license-based SaaS business as good and bad product experiences dont arent reflected in revenue during the period of the contract.
  • 03:59
    ROI delay in license based SaaS products
    In a license-based SaaS model, companies contract for a period, and decisions about product continuation occur at contract renewal. This poses a challenge for analysts to link daily user experiences to future renewal decisions.
  • 05:36
    Building prediction models to deal with ROI delay in license based SaaS products
    Understanding product utilization and its trajectory by tracking usage trends over time helps identify patterns like increased or decreased usage. These trendlines enable early detection of concerning signals and facilitate interventions to ensure customers derive optimal value
  • 07:35
    Strategies for extrapolating early usage signals into predictive insights
    Analysts and data scientists must be involved in early product discussions to define success criteria, translating them into trackable data points for accurate prediction models. Feeding data into models from the product's inception ensures that prediction models become more accurate over time.
  • 10:50
    Instrumentation driving decision making
    Instrumentation refers to translating things or actions we want to measure into data points. As more things are added to your instrumentation, you can get more rich information about the product experiences and help to drive the right kind of decision-making.
  • 13:09
    Driving customer value for long term financial success
    The goal is for your product to deliver value to customers beyond revenue. By using predictive models to understand and replicate successful customer patterns, you can drive overall customer success and long-term financial growth for your company.
  • 15:09
    Supporting SaaS users and customers for success
    Being customer-centric involves leveraging various channels, from improving the product to offering implementation assistance and user training, to ensure user and customer success. Its critical to utilize all of these avenues to drive success for users and customers.
  • 20:30
    Understanding the limitations of data accuracy for decision making
    Recognizing the limitations of data accuracy, it's essential to trust intuition, conduct experiments, and be patient with success while running a business. The aim is to offer "good enough" information for experimentation, measurement, and iterative improvement in decision-making.

Quotes

  • "The best company is a customer-first company. If you bring value to the customer, revenue is a byproduct of that." - Victoria Zhang

  • "Theres delayed ROI in a license-based SaaS business. As analysts, when we're thinking about bad user experiences today, we need to imagine the future and how this is going to impact long-term customer loyalty." - Victoria Zhang

  • "Prediction is a technique that will benefit from more data points. If we dont have instrumentation on a certain product feature from day one, we have even thinner data to feed into prediction models." - Victoria Zhang

  • "Instrumentation is a way to turn the real world into digital data points." - Victoria Zhang

  • "Sometimes we need to have good intuition to say, I've looked at the data, this gives me enough competence that we should test this method. If it doesn't work, let's shift the gear and do something else." - Victoria Zhang

About the speaker

Victoria Zhang

Salesforce

- Salesforce

Victoria is VP of Data Science at Salesforce

Up Next: