This Q&A with Infer’s Sean Zinsmeister was originally published on MarTech Advisor.
Predictive analytics for B2B sales and marketing has certainly “crossed the chasm”, but it’s still in the early adopters phase of the product development lifecycle, and will continue to mature. MarTech Advisor spoke to Infer’s Sean Zinsmeister about the predictive space and Infer’s product strategy.
Consolidation Versus Integration of Predictive Intelligence Platforms
I expect that we’ll see some consolidation, but even more integration as the industry evolves. The reality is that today’s marketing clouds are still fairly immature and cobbled together vs. providing one cohesive cloud. When you look at the predictive movement, there has been a lot of hype and a lot of vendors chasing shiny objects. Our strategy is not to build an all-in-one solution or a walled garden, but rather to deliver an open architecture that can share data, predictions, recommendations and action triggers across any marketing cloud, system of record or other specialized tool (i.e. AdRoll, Outreach, Pardot, Act-On, etc.).
Open architecture is especially important because martech is getting increasingly balkanized by salestech, and the go-to-market stack is expanding. Our approach is to build deeper hooks into engagement systems. This will in turn increase the predictive power of our models and allow us to drive more targeted segmentation, recommend appropriate next-best actions, and ultimately make all of a company’s systems run more efficiently.
Adoption of Predictive Intelligence by Enterprises and the Future of B2B Marketing
We are already seeing predictive analytics adopted by more enterprises, as shown by the recent Salesforce’s recent State of Marketing report where they found that 79% of high-performing teams currently use predictive intelligence. This is why larger vendors, like HubSpot, are entering the market with watered-down versions of predictive scoring, which is helping to broaden awareness for the opportunity.
However, the lowest common denominator approach also adds more noise, confusion and risk, which makes some B2B organizations skeptical about predictive as they try to figure out how it can improve their sales and marketing efforts. What they really need is actionable intelligence from the data vs. vast quantities of data that they don’t know how to operationalize.
At Infer, we see predictive scoring as a subset of a broader system of intelligence, which is the future of B2B Marketing. Predictive scores are just one criteria to bring into sales intelligence and rich, descriptive profiles of ideal customers. Profile management, on the other hand, is a foundational technology that helps businesses overcome the difficult challenge of effectively managing segmentation and targeting in a data-immersed world. Sophisticated profiling can help enterprises orchestrate their data by telling marketers who the most ideal target profile is for a particular campaign, and then sending the relevant list to the recommended engagement platform for action.
As the market evolves, it’s become clear that predictive scoring tools are not enough, and people want more transparency and control (for good reason). Across industries, machine learning outputs in general just aren’t that actionable on their own, and often end up as an incremental optimization vs. end state products. What our industry is missing is a guide that tells marketers and sales what to do next, vs. just giving them a score or a data entry system like Salesforce or Marketo—this is the gap we’re filling with the Infer Profile Management Platform.
Predictive intelligence for Predicting Intent and the Risks of Relying on data science and machine learning
When it comes to predicting a prospect’s propensity to buy, it’s important to keep in mind that there are two types of intent data signals (internal and external) – and they each produce very different insights. Activity from your own marketing automation system or application logs contains strong predictive signals and good coverage, especially when your behavior models are married with fit models in the right way. Although external intent data from third-party publishers like Forbes or TechTarget is an exciting new frontier, these signals simply don’t yet have enough coverage to be useful for predictive models.
In all of this, it’s definitely important to find a natural balance between human intuition and machine learning. While we have too much data in our world to process with the human brain alone, there’s a huge danger in relying blindly in machine learning – imagine how frustrating it would be if Waze took you two hours off your route because of some glitch in the algorithm. Data science is there to provide us with actionable intelligence and help increase the efficiency of programs, but it’s certainly not a silver bullet.
Using Account-Based Marketing to Take Advantage of Predictive Intelligence
The best way to think about ABM is to go back to the marketing mix, and consider where it fits as one part of your overall portfolio. ABM is really just a way for marketers to focus more on down-funnel accounts, but in order to close those accounts you’ll still need individual leads and contacts to target with your marketing programs. Many businesses adopt an ABM strategy in order to move up-market and capture bigger revenue pools where there’s a large density of high-potential buyers. Predictive intelligence can be a key part of pinpointing these segments to execute a sound ABM strategy, but it’s not a vital component, and vice versa. Predictive intelligence adds tremendous value to leads-focused marketing strategies as well.
ABM can play a big role in all types of B2B businesses – even those with a fairly large universe of accounts. For example, our customer InsightSquared uses predictive intelligence to manage ABM at scale and answer the questions: “How effective are my ABM programs? Are they helping to accelerate pipeline?”
A major benefit of predictive technology can be sifting through mountains of data to help you gain a clearer, crisper understanding of who your buyer is, and this is a foundational element of any good marketing strategy. Sophisticated profiling and segmentation have a major impact on marketing effectiveness. We’ve covered several high-impact use cases including architecting highly effective nurture programs, accelerating expansion, powering personalization at scale, measuring the effectiveness of marketing campaigns, etc.
In particular, our customer Social Tables used predictive to uncover new revenue and expansion opportunities that increased its opportunity pipeline by $500k and overall revenue by 7%.
Infer offerings and Strategies for Product Development
Infer stands out because we were the first company to truly productize predictive scoring, and we’ve steadfastly focused on building a successful community of customers who are now consistently achieving real business value from predictive. Very few of the vendors in this space are demonstrating customer success stories with tangible results like this. Many seem to be too focused on consultative approaches that require lots of professional services, or theoretical approaches that over-promise and under-deliver.
At Infer, we’re focused on three key pillars of innovation that we believe will redefine the sales and marketing stack. In order to evolve the basic data entry platforms that are today’s CRM and MA systems of records, we’re focusing our R&D efforts around building a “system of intelligence.” It leverages an open architecture to wrap data science around today’s stack via predictive scoring, profile management and sales intelligence. These three legs of the stool are interconnected and equally important in helping companies move beyond the basics of predictive scoring.
Infer Glance is an example of our product innovation in the sales and account intelligence area, and it’s something that we developed in beta for about two quarters as we validated it with customers and fine-tuned the solution to deliver real impact for frontline sales reps.
Infer’s Technology Challenges and Innovations
As we’ve scaled our technology one of the biggest hurdles was opening up our platform for any company to be able to try out in a free-trial program – something we were thrilled to launch last month. While this step may seem simple from the outside, when it comes to predictive technology, it’s actually something that requires a lot of work behind the scenes. This milestone was key in allowing us to get our product into the hands of more influencers and early adopters, so that we can find more ways to map our solutions to real-world problems.
Another challenge we face is determining which engagement platforms will be in the most demand beyond big systems like Marketo, Salesforce and Eloqua. New technologies are constantly emerging in the sales and marketing landscape, and we want to infuse predictive value across a variety of go-to-market workflows.
On the horizon, I’m excited about how we’re infusing more and more data science into automation systems to finally deliver on the failed promise of marketing automation, which was to bring actionable intelligence to marketing teams.
With Infer’s intelligent recommendations, our mission is to make it dramatically easier for sales and marketing teams to adopt predictive into day-to-day workflows, identify ideal customer profiles, and determine next-best actions for engagement. The new capabilities we’re working on in this space will go a long way towards advancing our mission of helping businesses grow through the power of data science.