We wanted to write about our learnings from the past few weeks as we explore the right product-market fit for a solution to help customers with customer and prospect contactability. This blog explores how the data that we provide via our Phone Lookup APIs is being used by our customers and whether there are ways we can solve some of those business problems ourselves rather than just providing raw data and letting our customers figure out how to make the most out of it. Our goal for our contactability solutions is to democratize them with our data/insights expertise so they can be leveraged by big/small companies across the board.
As we talk to our customers, especially in the lead management space, two things always come up in terms of how our data is being used:
- We want to weed out the worst leads before we provide them to our clients.
- We are resource constrained, and we want to prioritize the best leads to call or reach out to.
In industry terms, both these problems are referred to as ‘contactability’ of the leads. Essentially, businesses want to know whether the prospect/consumer can be ‘contacted’ with the information provided. This is especially true in today’s world, where the pick-up rates for incoming calls are at an all-time low, especially when the number is not recognized. We found that according to Pew Research, 80% of Americans in 2020 and 87% in 2022 according to this article do not answer unknown incoming calls – a steep increase for an already challenging statistic. Hence, targeting the right audience who are most interested in your solution rather than wasting time on leads with incorrect or junk data is more important than ever.
As an aside, this is also important for outbound callers who do not want to dial disconnected numbers or numbers that do not belong to the expected customer. There are multiple reasons for this including regulations and avoiding being marked as spam by carriers. We will cover this business problem and how Trestle can help in more detail in another blog post.
Maximizing Contactability: Key Insights from Analyzing Lead Data for Weeding Out Junk and Prioritizing Good Leads
Coming back to weeding out the worst leads and prioritizing the ‘good’ ones, here are some of the things we are learning, both from talking to customers and collaborating with them on in-depth data analysis:
- Leads that have invalid phone numbers or are disconnected are obviously junk and should be immediately discarded.
- Phone line types that are not mobile have very low contactability, especially if they are non-fixed VOIP phones like Twilio or Google Voice. Of course, there are exceptions.
- Leads where the name provided on the lead form matches with Trestle’s phone data is a great positive indicator of contactability.
- Finally, any kind of activity seen on that phone, especially if it’s more recent, is another positive indicator and increases the contact probability.
Building the Perfect API Product: Balancing Feedback from Wide-Ranging Customers and Channel Partners for Optimal Contactability
We are gathering feedback on what our product should look like. For example, here are some of the questions we are trying to answer: what are the right inputs, and what are the right output/response attributes if we bring an API product that answers the contactability question about the lead?
The challenge we face is we are focused on a wide spectrum of target customers, both big and small, and our customer engagement approach is working directly with customers and channel partners. For example, a channel partner wants the product to be as self-serve as possible, as neither might have the bandwidth nor expertise to learn about all our raw data attributes and figure out how to make sense of it. They want to be able to implement a simple rule using one attribute (think score or grade) from our API response, and that should suffice to either weed out the bad leads or prioritize the good ones with sufficient accuracy. In other words, that one attribute threshold should be highly predictive (high correlations with good/bad leads) and works for most customers.
We also hear from our more analytical customers that we should provide as much data as possible so that they can put that in a model that covers their use case-specific nuances to optimize for lead quality and quantity. Was it a valid/invalid phone, what was the line type, what/how much activity has been seen with the phone, did the name match? – these are all attributes that go into the score/grade, and making it available individually helps sophisticated customers who want to leverage the raw data to feed their models.
As I mentioned at the beginning, we are still in the learning phase and talking to as many customers as possible. We’re working with them to analyze their data with the outcomes (good/bad lead) and correlating it with all the data we have. There are still many questions we are trying to answer: What scores/grades, which data signals to provide, how many attributes are too many, etc.? How would we position this in the market, and what is our right to win against our competitors? Solving questions so that we can align the business problem, the solution, its packaging and pricing for the customer engagement channels targeted, all with appropriate sales motions and customer service – that’s what gets us excited. And we would love to have you as part of this journey.
If you are one of our customers or are looking to explore our data for lead management solutions, we would love to talk to you and learn more about what the right product for you looks like.