In 2016, Datavail, Colorado-based provider of data integration and database administration in North America, began to revamp its marketing strategy to improve demand waterfall conversion rates and marketing campaign performance. Delivering data services for over a decade, the company is currently providing business intelligence, application development, and database support and services to more than 300 customers, boasting an average seven-year client retention.
Datavail’s goals were to increase demand generation waterfall effectiveness by identifying leads and accounts with the highest propensity to buy, and drive more pipeline value by identifying leads and accounts with highest tendency toward larger deal sizes. The company also wanted to identify new accounts that look most like their customer base and are in-market to optimize account-based marketing (ABM) efforts using data-driven, strategic methodology, and increase marketing-attributed revenue. Megan Isherwood, the company’s Director of Marketing Operations, was sure they could get there by leveraging predictive analytics.
Leading the implementation of predictive analytics at Datavail, Megan, whose career experience spans marketing, demand generation, and operations, explained that her team is measured by marketing initiatives that generated closed business—not just influence. Being responsible for 75% of the funnel, the team is charged with a huge target and has a monumental influence over the company’s pipeline and financial results.
Saying no to black box
During Datavail’s search for predictive analytics vendors, Megan shared how she was turned off by vendors with a “black box” approach. “One of our key drivers for signing up with Mintigo was being able to log in to the platform and see any indicator that we wanted,” she said. “We can get more and deeper data than what was readily available to us.”
Creating goal-specific predictive models
Megan and her team operationalized predictive analytics in lead scoring and data enrichment in their marketing automation and CRM systems to enhance demand generation programs across multiple channels. Before attaining remarkable results, Megan conducted several critical activities and tests to ensure the successful implementation and integration of predictive analytics into Datavail’s existing processes.
To align with Datavail’s services, three different predictive models and use cases were created. Megan and her team came up with both account-based and lead-based models for remote database services, and an account-based model for analytics services. Each model with its own target was connected to relevant departments within the company.
For instance, the lead-based model for remote database services ranked new leads that entered the funnel, and prioritized inbound leads for sales follow-up. This model leveraged Mintigo ranking and specific marketing indicators for those type efforts. Meanwhile, the account-based model for analytics services considered company growth, which required development of product-specific models. Hence, the model identified existing accounts to target for cross-selling new products, on top of net new accounts and leads more inclined to purchase new products.
Validating and measuring before implementation
After the predictive models were created, Megan initiated the validation of program results to ensure model performance accuracy. “We built the models privately before we rolled into our sales team so that we can measure without implementing the human behavior. We used different funnel tools to validate that data was sitting in the right places before we start adjusting our strategy based on A, B, C, D lead scoring. We wanted to be 100 percent sure that it was working,” noted Megan. Doing this enabled her and her team to track across stages and measure results more easily and accurately.
After monitoring and validating the model performance for a quarter, Megan confirmed that they can influence how many of the leads move across the funnel by triaging using Mintigo models. “Another validation point was that As and Bs, once they move forward into the opportunity stage, closed almost twice as often as Cs and Ds.”
Seeing results in all three use cases
One of the marketing programs launched by predictive analytics was their custom board game Datavailopoly, a high value direct mail offer about Datavail services. Megan and her team launched the program in four phases: two before predictive marketing and two after.
Before predictive, the Datavailopoly program generated 131 percent ROI with zero pipeline remaining. For the second phase, they used predictive modeling to drive list selection for executives that they planned to approach for the campaign. After predictive, not only did the program drive 146 percent ROI, but it also generated a significant volume of late-stage opportunities that had the potential to increase ROI to 1,700% if successfully closed.
“Really, the only change we made was using predictive for the list selection and prioritization. We’re pretty pleased with the results knowing that the dollars generated is our monthly reoccurring revenue, which means customers will stay with us for the next eight years,” Megan shared.
Another use case–display advertising–utilized predictive modeling to target an A and B list for database services through search. The campaign resulted in an uplift in conversions across all display advertising by nearly 40%. As the graph shows, there was an impressive uplift of 106.6% in click-through rate (CTR) and 168% in conversion priority displays, which were the best performing ads with people Datavail wanted to reach. The online search programs also generated more clicks preceding the conversion point with no additional spend or messaging changes.
The last program was a multitouch nurture campaign called job posting, where people looking for database jobs to be filled were contacted and pitched database managed services. With predictive analytics, Megan led business development representatives’ (BDR) follow-ups by eliminating the Ds and sorted the list so As were on top. The result was a significant upsurge in ROI from 72% to 265%.
Launching predictive, Datavail used the Q2 2016 baseline with 72% of bookings generated by A/B accounts. In less than a year, Datavail’s Q1 2017 A/B account bookings were 329% of Q2 2016 and Q1 2017 overall bookings were 206% of Q2 2016. Megan and her team were pleased to see these promising results as higher value customers stay longer and are more promising for cross-sell and upsell opportunities.
Implementing any new technology or process in an organization is always difficult as some individuals are change averse. When asked how other teams reacted to the results of predictive analytics at Datavail, Megan explained, “They’ve been thrilled every time we roll out to a new team especially since we have these analytics to back up what we’re saying; we don’t need them to just take our word for it. They are excited and have immediately adopted Mintigo. Once the BDR team was on board, we started rolling out to test teams, then our account representatives.”
Moreover, results derived from predictive models also served as a training mechanism. Megan explained that at Datavail, there were sometimes junior and senior account representatives in one region. When predictive and lead triage were rolled out, junior representatives took the C and D meetings while senior representatives took A and B meetings.
Seeing the future with Mintigo’s predictive analytics
As they continue to embark on the journey with predictive analytics, Datavail is planning to build models for newly acquired lines of business as they need audiences that are currently not in the company’s existing database. They also plan to roll out predictive scoring to all sales and BDR teams, apply analytics to a new strategic account sales planning process, and explore catering to multi-service type models for cross-selling or upselling to existing and new accounts.
Finally, Megan and her team have set their sights on taking over the model building process internally, which would eventually be more productive. Datavail’s experience in implementing predictive analytics emphasized the need for having concrete goals, validating models, and measuring results not only to monitor performance but also to easily encourage the rest of the organization to use predictive analytics.