COVID-19 Update

Why Infusionsoft Evolved To Predictive 2.0

  • Written by Doug Sechrist, Infusionsoft
  • Published in Demanding Views

Doug Sechrist 1As a B2B marketer, my mission is to drive revenue.

This mission is one that I’ve been privileged to take on time and time again to multiply the customer base at several of the top enterprise software companies on this planet. From early days at Eloqua to Five9 and now Infusionsoft, the pipeline and leadership expectations for the marketing team has always been huge, requiring a vision and team that will hopefully act as a precedent for other marketers faced with aggressive goals. While I’ve had some early wins and lessons learned – I am now tackling the steepest challenge of my career as the VP Demand Marketing at Infusionsoft.

The mission: Lead the movement to raise the success rate of small businesses worldwide.

How will we get there? It starts with understanding what it takes for the millions of small-to-mid-sized businesses to be successful and developing tools which enable that success. Next, we have to reach, educate, and empower all of our future customers down the path to growth. That’s where my team comes in, and it manifests in 10x targets for revenue and customer growth metrics by 2018.


No goal in my career, big or small, has ever been achieved without the help of technology. Today, we’re lucky to have constructed an integrated technology stack and process that works as a demand generation factory that is consistent and predictable. But at the outset, in order to dominate Infusionsoft’s massive market of all-in-one sales and marketing software for small businesses, I, like so many other modern marketers, saw the promise of predictive align perfectly with my needs.

However, my first experiences with early-generation predictive would not fulfill the needs for Infusionsoft. Let’s take a closer look at my journey, the lessons learned, and a visual of the technology stack working for us today.

Lessons Learned From Predictive 1.0

My hope is that before you invest in a solution, you can skip a few steps and mitigate many risks by learning from my experiences. My experiences are not unique. I travel to speak at many conferences and have heard these same challenges echoed from many other innovative marketers that stretch the boundaries using new technologies like predictive.

Scalability of insights across the buyer journey, inability to power outbound, and output quality due to data coverage and accuracy were the three most commonly referenced limiting factors.


When I started at Infusionsoft in mid-2015, I knew the funnel needed rebuilding. A newly minted funnel structure would give us a better understanding which levers would drive acquisition and would allow us to add more levers (aka channels).

In a dynamic, large SMB market, accessing the data and insights across a buyer’s journey was only mildly achievable with custom-built predictive solutions that modeled data for various scenarios. For example, if 20% of our likely buyers for next quarter were currently in the early inquiry stages of the funnel, my team should be shifting resources toward outbound activities like custom nurtures, high-touch telemarketing, and content centered around ‘how-to’ and ‘why now’ messaging. With early predictive solutions, each scenario required building a new model, which doubled our cost and required multiple months before a go-live date.

Missing Outbound

The majority of Predictive 1.0 vendors began with lead scoring solutions that helped solve the inbound marketing dilemma of “too many leads.”

But scoring inbound is just one component of building the demand factory, especially when we could use outbound to pursue the right accounts rather than wait for SEO and paid channels to attract them. Without our predictive solution sourcing net-new leads, we had to add data providers and self-prospecting sales rep alongside scoring vendors to stack rank for the sales team. Operationally, it worked – but it wasn’t streamlined, data sources weren’t reliable and we were adding to the unhealthiness of our CRM.  We also relied on our marketing automation technology (MAT) to segment audiences that would be used for improved outbound programs. While great for behavioral segmentation, MAT relies 100% on your existing data. Even with fill rates across basic fields outperforming industry benchmarks, our data would not leverage all the available data that could help us identify our target market.

The current process from data to campaign launch added more providers, costs, and complexity to the workflow — which was in direct opposition to my goal of “constructing an integrated technology stack and process that works as a demand generation factory.”

Garbage In, Garbage Out

It always seems to come back to data quality and coverage. The outputs of predictive – the scores, target accounts, micro-segments – all rely on the external data inputs used by predictive platforms. The better the inputs, the better the outputs.

While the promises from predictive vendors are exciting, buyers must be confident in the data inputs before we take action or set forecasts. Things like coverage in the target market, data validation and update methods, and accuracy benchmarks are important to understand before taking predictive outputs at face value. For Infusionsoft, if there wasn’t coverage and accuracy in the SMB data, then no predictive use case or interface would ever work.

At the end of the day, I needed to start small with a specific use case – enable better outbound. I used the gaps we saw with other predictive solutions to assess new options.

How Predictive 2.0 Completed Infusionsoft’s Demand Factory

The additions that were critical to evolving Infusionsoft’s demand factory were simple:

  • Add new acquisition channels leveraging outbound lead generation 
  • Offer high-value content aligned to the customer journey to drive engagement

A new solution would have to offer new prospect sourcing within the target market and robust segmentation features as core functionality. It would also need to eliminate campaign execution steps and vendors by coupling application capabilities with an accurate database of business information with coverage across all small businesses. Simplification was a must.

After assessing several solutions (including one legacy solution we had already implemented), we found Radius to be the best fit for our needs. It integrated with our current ecosystem and offered analytics and data for faster time to deployment of multiple models, accuracy and breadth of data used for both predictions and outbound campaign execution, and a user experience that was very intuitive. Once implemented, it would allow our team to be self-sufficient in driving demand throughout the re-architected funnel.


Now equipped with a new predictive solution, our team is full-speed ahead on our mission. We’re applying Radius for four specific use cases: 

  1. Refine Ideal Customer Profiles (ICPs): Use analytics powered by a rich SMB dataset to obtain a 360-degree picture of what makes our best customers unique. ICPs are used across content strategy, personas, database segmentation, and net-new sourcing
  2. Score existing database: Apply scores to every lead as an indicator for how close a prospect is to the ICP. Setting priorities for both marketing and sales has accelerated the existing pipeline.
  3. Layer outbound segmentation: Slice and dice existing database by likelihood to convert and add to email nurture programs for education down the path to purchase. Over time, the segmentation has become more granular, resulting in multiple nurture streams based on top predictive signals and relevant content offers.
  4. Inject outbound outreach: Target lookalike accounts that match our ICP, which has grown outbound outreach by 200,000 prospects per month.

Our dual-threat of inbound and outbound ensures no lead is left behind. Scoring and nurtures that marry content with prospects ensure inbound leads have consistent education and cultivation, and when the moment is right, hit a threshold for sales outreach. Net-new, segmentation, and integrations into CRM ensure that we are reaching all of the businesses that are ideal fits for Infusionsoft.

With the right predictive solution acting as a backbone system of insight, we can target prospects that match our Ideal Customer Profile. We see Radius as eHarmony plus Spotify for our CRM. Our mission has just begun, and we will continue experimenting and deepening our usage of this technology to drive revenue at InfusionSoft and help every small business succeed.