When I was running enterprise marketing for edX we evaluated ABM software vendors and ultimately selected 6sense (read my post about evaluating ABM vendors here). I thought I would publish a post about what it’s like to deploy ABM, configure the ABM platform, and roll it out to the sales and marketing team.

1) Configuration

It took about three months to configure 6sense, build the AI model, integrate with adjacent technology (Salesforce, Hubspot, Google Analytics), and roll it out to sales & marketing. You’ll need to work closely with operations to get a data dump of several years of closed sales data — that’s a major input for tuning the AI model — and to configure the integrations.

2) ABM Model Creation Strategy

What’s your strategy for ABM model creation? This is something you should discuss with your ABM vendor. Should your AI training data set include all sales data, for an unbiased representation of your historic sales? Doing this can be very informative for debunking your Ideal Customer Profile (ICP) and your overall go-to-market (GTM) strategy. Do your ICP definition and your GTM strategy really accurately reflect the actual sales that have closed in the past several years?

When we looked at our real unmanipulated data it showed that we tended to sell reactively to inbound leads, selling a lot of small deals to SMB businesses, even though our aspiration was to sell proactively, landing bigger deals to larger companies. During AI model creation you have the opportunity to game the system by providing a training data set that represents your aspiration rather than your historical reality.

This is a good conversation to have during ABM vendor evaluation. Some ABM vendors were very rigid about not manipulating the ABM model, treating it like a sacred cow. We felt it was critical to have ABM lead us to sell differently and it defeated the purpose to build a model that would become a self-fulfilling prophecy for the same old way we were already selling. We chose an ABM vendor that would help us achieve our aspiration.

3) Sales Pilot

Once the integration & model building were complete, we rolled it out to a pilot group of 4 sellers. The pilot phase is critical for working out bugs and making sure the model is identifying good fit prospects that represent your ICP. It also gives you the chance to train sales and marketing users in the system and make adjustments to those curriculums.

An important biproduct of the pilot is that you start to grow buzz with the sales team. Pilot participants (hopefully) start to get excited about the new opportunities being delivered to them and word will spread to the other sellers.

We expected the pilot phase to take a month but it went smoothly and after 2 weeks we rolled it out to the whole edX enterprise sales team.

4) How ABM Changes the Salesperson’s Daily Life

ABM software is meant to transform the sales process. The first thing a sales rep should do each day is review ABM dashboards and alerts. The sales rep is expected to spend most of their time and effort pursuing accounts with the highest “in-market” scores — these are accounts the AI has determined are most likely to close quickly. Sales leadership should be aligned and should manage their team’s pipeline and activity accordingly.

Don’t underestimate how hard it is to change this behavior. Don’t cut corners on training for the sales reps and sales leaders, and user support for problems & questions that pop up. Monitor sales rep usage of ABM and responsiveness to alerts. Review ABM reports in your deal desk meetings to ensure in-market accounts are being worked and prioritized.

Don’t let sales waver and relapse to old ways of doing things when the end of the quarter comes and sellers are scrambling to hit their numbers.