Planning for 2014 – Why Predictive Marketing Should Be At The Top Of Every #MKTGnerd’s To-Do List

Now is the time to start planning your marketing strategy and programs for 2014.

Most marketing nerds are in the middle of their planning and budgeting exercises for next year. Allow me to make the case for why predictive marketing should be at the top of your to-do list for 2014.

By using all the data in the world, you can optimize your revenue funnel to simultaneously improve conversion rates, increase revenue and improve lead velocity. Predictive marketing amplifies your measurable marketing initiatives and provides an added level of performance transparency not achievable in the past.

WHAT IS PREDICTIVE MARKETING?

Predictive marketing works by taking all the data in the world, from internal (CRM, Marketing Automation, etc.) and external (blog, websites, government sites, social media, etc.) sources and applying modern data science to the questions:

  • Who is going to be my next customer?
  • How do I convert them?

Predictive scoring models are 3-10X more accurate that the traditional point-based scoring used at most organizations. Think about that – less work, more accuracy!! For the remainder of this blog post, I’d like to tackle the questions I frequently hear about predictive marketing from B2B marketers.

HOW DOES IT HELP MY MARKETING AUTOMATION EFFORTS?

Now that you have set up your revenue waterfall, it is time to optimize across the funnel. The details vary, but there are typically three elements to the solution:

  • Feed the front of the funnel with the best targets,
  • Send only the leads most likely to convert to sales while nurturing those not yet ready, and
  • Align sales activity to opportunity so that they spend most of their time where the expected value is the highest.

Each one of these solutions can be deployed independently or in concert.

HOW DOES IT WORK?

Predictive marketing models start with data that you already have in your CRM and Marketing Automation systems. This is enriched with hundreds of buying signals mined from the Web and third party data sources. By training on recent transactions or closed-wons, the patterns underlying your business are discovered. When we talk about Predictive Lead Scoring, the patterns are converted to a probability and expected value score for each account and each contact.

All you need to do is provide us API access to your CRM and Marketing Automation system and the scores will be calculated automatically at a frequency set by you. Over time, you can continue to expand use-cases and improve the scoring by adding more data sources, e.g.., product usage, service logs, renewals database, etc..

HOW MUCH IMPACT SHOULD I EXPECT?

You should expect a minimum of 10-15 percent improvement in revenue realization by optimizing any part of the funnel (TOFU, MQL, SAL). By optimizing it simultaneously across the funnel you will be able to increase revenue attainment by 20-25 percent.

WILL THIS BE A LOT OF WORK FOR MY TEAM?

If you are using standard cloud-based CRM and Marketing Automation systems, we will need just 2 hours from your team to get this launched.

DO I NEED CLEAN DATA TO GET STARTED?

No, you don’t. Our applications rely heavily on external data, which is cleaned by checking for consistency across sources, it can work with imperfect data on your end. Given any company name or individual email, the application can automatically identify hundreds of buying signals for the company, disambiguating where necessary between headquarters, business units and sites.

WILL THIS WORK IF I DON’T HAVE A TON OF RECENT HISTORY?

Predictive scoring models can work with as few as 25 initial events. So if you have more than 25 customers, you are good to go.

CAN I DO THIS IN-HOUSE, BY HIRING A FEW DATA SCIENTISTS?

If your use-case relies on scoring based on internal data only (e.g.., predicting renewals), an internal effort can get you part of the way there. However, for most sales and marketing efforts, external data and dynamic buying signals tend to be as important in predicting purchase. By plugging into a cloud-scale infrastructure to collect and interpret this data, you can avoid the cost of an expensive build-out of a data-collection system.  Since our applications start at a few thousand dollars per month, the ROI relative to building a data science team from scratch is very compelling.

WHAT SHOULD I LOOK FOR IN A PARTNER/VENDOR?

You should look for eight things:

  1. Domain expertise – Does the provider understand marketing and sales or is it a technology looking for a problem?
  2. User-acceptance – Does the provider have tens of thousands of end-users who have accepted their recommendations or should you expect push-back from sales once deployed?
  3. Data Completeness – Are they using internal and external data or internal data only?
  4. Scale – Do they have experience with large data sets and modeling at scale vs. deployment at small companies only? Will they be able to grow with you as you grow?
  5. Scope – Can they cover a sufficiently large number of use cases so you don’t have to buy a new system for each use case and stage of your funnel?
  6. Security – Do they support enterprise-class security? Do they have customers who are as serious about security as you are?
  7. Time to deploy – Predictive scoring applications should launch in days not weeks. Is the provider committed to a quick deployment or are they after large services and implementation fees?
  8. Stability – Is the company stable? Does it have at least 50 customers, so you are not funding someone’s science experiment?

FINAL THOUGHTS

A predictive approach to scoring at the account and contact level is one of the few revenue generating initiatives that can have an impact in days instead of weeks. By its very nature, it is measurable and quantifiable and works within your current workflow. It will help your organization predict your next customer and convert them in a measurable, scientific way. Done right, it will make you a revenue superstar in 2014.

Written by

Shashi Upadhyay
October 18, 2013

Comments