“What elements need to be in place before starting a predictive marketing program?”
We get asked this question a lot. Predictive marketing is all about good data quality, so a better question is,
“How do I make sure my marketing processes and technology generate high quality data?”
Graphic: Key sales funnel actions and data
Here is a checklist of the system capabilities and operational principles that can help you improve the quality of your data when setting up a predictive marketing system:
- Capture interaction data from all touchpoints. Predictive marketing data needs to reflect the actual customer journey. Almost every customer’s journey is unique, so you need to collect transactional data from each and every interaction with a customer. This is easy in the digital world, but you should also put in place mechanisms to capture IRL (in-real-life) events like store visits and encounters with potential customers at trade shows.
- Tie interactions to a contact. In direct marketing, without an identity–an actual person–there is no one to market to. It’s easy once a contact has submitted a webform, but you also need to be able to capture anonymous interactions and later tie them to an identity. This is typically done by placing a cookie in their browser. Then, once they submit a webform, the cookie and all related interactions can be tied to that contact. Keep in mind that this can get quite complicated when contacts use multiple devices, sometimes even multiple browsers, from many locations with different IP addresses.
- Assign contacts to a company. If you cannot assign a contact to a company, their value to you as a B2B customer is very limited. Once a contact submits a webform with their company name, this task is quite easy. But you can and should match contacts to their company before knowing exactly who they are: find out their IP address and then use reverse-IP lookup to map them to a company. This is especially important in Account-Based Marketing (ABM) where not all buying team members submit a webform.
- Maintain and enrich master data. Keep your more static contact and company data in good shape (up-to-date, accurate, no duplicates, etc.) and enrich it with information that can be used in market segmentation and lead prioritisation. This includes contact data like job titles and firmographic data like industry classification and company size.
- Tag marketing channels like ads, referral links, and search words with UTM parameters. There are four touchpoints in the customer journey that you absolutely need to capture: first touch, lead capturing, opportunity creation and deal closing. If you can pinpoint the marketing or sales action that preceded each of these conversion points, you already have a good game plan for converting your next customer.
Once customers enter your digital domain (Marketing Automation Platform and CRM = the last three touchpoints) this becomes fairly simple, but you can only analyse the effectiveness of the first touchpoint by tagging all your marketing channels and campaigns (display ads, social media, referral web pages, third-party apps etc.) with UTM parameters.
- Do not manually touch or manipulate your data. This is very important. Unless you are the data scientist responsible for ETL (and if you don’t know what this means, this is probably not you), do not aggregate data or transform the data in any way. Don’t manipulate the data by hand in Excel. Data extraction and transformation must be done using scripts, because manual handling of the data almost always reduces its quality. Either give your data scientist direct access to the database or deliver the data in its raw format.
- Prioritise your customers. Some are simply more important to you than others. In all predictive marketing processes, it’s important to ascertain each customer’s relevance to your business. It’s essential to know three key pieces of information for each customer: 1. total revenue during the past year, 2. the number of sales transactions during that year, and 3. the date of their latest purchase.
While this might sound like a lot of work, your company is most likely already doing most of these things. You simply need to make contacts with the right people in your company in order to get access to this data and make sure the principles listed above are being followed.
There is no better way to get to know your customers than predictive marketing. And those companies that know their customers the best will win the game in the long run.
About the writer:
Matti Airas is a consultant in predictive data-driven marketing and customer experience. He has previously worked for the customer experience feedback analysis company Etuma and for Nokia in the U.S. His passion is figuring out how to use data to solve business problems.
Matti enjoys writing, podcasts (especially on U.S. politics), golf, long walks with his wife and Jack Russell Terrier, and any kind of skiing.