Implementing Predictive Intelligence Part 3: Where Predictive Fits In
In the third installment of our Implementing Predictive Intelligence series, we wrap up our look at Cisco’s approach to and learnings from rolling out auto-qualified leads. Our previous posts explored the launch of Cisco’s predictive intelligence initiative and provided 4 recommendations to ensure the success and stickiness of your predictive efforts.
This post will examine four key areas to consider as you plan for predictive intelligence: where to apply predictive, your organization’s capacity for change, data hygiene and a phased rollout.
1. Where to implement predictive
For Cisco, a major benefit of rolling out predictive is gaining visibility beyond their own ‘storefront.’ Unlike their pre-existing scoring models, 6sense’s predictive intelligence platform taps into data external to Cisco.com to give their team insights into their buyers’ broader digital journey.
Deciding how to apply these insights into pre-existing workflows is a key decision that is unique to every organization implementing a data-driven approach to marketing and sales. For Cisco, top-scored prospects go directly to sales, bypassing a lengthy and expensive pre-qualification process. Prospects who are earlier in the buying process enter a nurture flow that includes personalized emails and website customization highlighting the technologies and solutions they demonstrated interest in.
Establishing these paths required significant planning. For Sean Beierly, who oversees the program for Cisco, the linchpin of the process was building credibility for predictive results before attempting a pilot with direct-to-sales leads.
2. Know your capacity for change
Understanding your organization’s tolerance for change and openness to embracing a digital transformation should guide your predictive intelligence implementation. Having an internal champion and executive-level buy-in will ensure the program gets adequate support to succeed.
As Cisco rolled out predictive intelligence, the majority of demand generation activity continued unchanged. Their aim was to improve existing processes first and think of replacing those once new methods were tested and proven at scale.
Their initial focus was to implement predictive with the sales teams who had bought into predictive intelligence and had the bandwidth to collaborate on the implementation. Together they would test and fine tune the models, rather than trying to enlist the support of skeptical or thinly spread teams who didn’t have the extra resources to drive a successful outcome for the initiative.
3. Data hygiene and lead flow
Ensuring that your predictive initiative is in full compliance with your privacy and security processes is an important step to getting the right data into the hands of the teams that need it. Equally as important is a clear understanding of the suppression and de-duplication processes to not only enable the right leads to get to marketing and sales, but also to protect your customers’ experience. The last thing you’d want is for predictive intelligence to degrade, rather than improve, your prospects’ engagement with your brand.
Predictive intelligence identifies in-market prospects — accounts and contacts that are exhibiting strong buying signals. Because these insights are time-sensitive, it’s important that marketing and sales teams take action immediately. So the speed at which these insights are delivered and acted upon will greatly influence the measurable outcomes of predictive intelligence implementations. This step pairs well with the lead flow process exercise we described in our first post outlining Cisco’s predictive intelligence efforts.
4. A phased rollout
When you are introducing something new, a phased rollout makes a lot of sense. For Sean’s team, this approach reduced complexity, kept the workload manageable and created a feedback loop that served to improve their predictive models.
The team focused on the US region first, which allowed them to work on one model, one set of business and data requirements and one group of stakeholders. A phased rollout also allowed them to continuously iterate on their model and improve the signal to noise ratio.
Cisco focused their predictive modeling on a single product amongst the many product lines and solutions they offer. This approach allowed them to zero in on foundational concepts that would pave the way for success in other regions. They were able to monitor the current work process of their marketing and sales teams, improving results without disrupting existing processes. Finally, they implemented a feedback loop to understand how to tweak and improve the models over time.
In our next installment, with some help from Kerry Cunningham of SiriusDecisions, we’ll discuss the risks and rewards of implementing predictive intelligence to your sales development and qualification teams. In the meantime, to learn more about how to implement and roll out predictive intelligence connect with our team and save a seat at our upcoming webinar.