Why do data integrations keep marketers awake at night? In all the recent marketing research I found, ‘data integration’ was named as the top challenge of marketing departments. Marketing platforms integrate pretty easily, data flows both ways. So, what’s the real issue?
Well, let’s start with the phrase ‘data integration’. Does that mean the technical knowledge required for integration? Having the right data? Or is the biggest obstacle for B2B marketers simply a massive disconnect between ‘data integration’ and improving marketing > sales > revenue?
The amazing difference between Data and Knowledge
The concept of ‘data’ is itself problematic. It’s a term that leaves me cold. ‘Data’ has not inspired me to higher win rates, larger deals or better margins. Even the word cannot be less appealing.
Knowledge is the opposite. It has charm. It suggests power.
Customer knowledge leads me closer to deals and long-lasting business relationships. Market knowledge helps me to navigate towards oceans of opportunities. Competitor knowledge guarantees me the right punch lines in face-to-face meetings.
Data stops the show. You can bang your head against the wall all you want with Big Data—gigatons of bits and bytes are not going to help you sell.
Is the word ‘data’ misleading us? Deceiving us? Yes.
I experimented by replacing ‘data’ with ‘knowledge’. Immediately, all jargon started to make sense. See for yourself:
Combine first- and third-party data sources to gain a more complete view of your audience.
…and the same sentence with knowledge treatment:
Combine first- and third-party knowledge sources to gain a more complete view of your audience.
See what I mean? Here’s another one where I replaced all ‘data’ with ‘knowledge’:
Customer knowledge refers to all personal, behavioural, and demographic knowledge that is collected by marketing companies and departments from their customer base. The collection of knowledge is aimed at finding insights into customer behaviour and maximizing profits.
Know why this works? Because data is like a mineral. Most of it is useless stone, but there are valuable particles that we can refine into knowledge. To do that, you have to know where to dig, how to refine, purify—and of course how to use. Once you have the knowledge, you have power. You can do something with it.
So, how do you convert customer data into customer knowledge?
In order to win each micro-moment of the journey, marketers need to deliver better, more relevant experiences—NOT FROM YOUR VIEW—but from the customer’s. Understanding motivation at each phase of the journey is the golden bridge you must build. And the bridge’s building blocks are alluring, brass-tacks, amusing or otherwise attractive content. It can range from Twitter comments to videos to eBooks, heavy-duty white papers or research documents.
Do you know what prospects are thinking?
The following diagram will give you some general guidelines.
The simple, universal fact is that prospects are not keen on getting offers in the research phase. You can accelerate the velocity of the purchasing process, but you cannot skip phases.
You have to tailor the questions in the above diagram to fit your line of solutions and vary them to capture the essence of each issue from the buyer persona’s point of view.
So, what’s next?
By finalising the journey story from the buyer’s angle, you can fill in the gaps of missing knowledge.
Computing power has gone up and the cost of storage down, so there’s really no reason not to store every interaction a customer has with you.
The types of knowledge you need
You can slice and dice prospect/customer knowledge in many ways, but the following six categories work well for most purposes. These are the kinds of knowledge that are useful to marketers, especially with respect to setting up a marketing automation program:
- Target company knowledge: In addition to firmographics: what are their needs and challenges, what’s their focus, who are their key competitors?
- Channel knowledge: What search terms are most popular; what are the hot forum and opinion-leader topics? Where are the hot spots where my audiences spend their time?
- Content knowledge: Of the topics, you identified in the previous step, what gets consumed? Is it eBooks, videos, case studies, comparisons, reviews & ratings, etc?
- Engagement knowledge: To what extent do prospects/customers engage and by what means? What are the results? Are they filling out surveys, adding ratings, leaving comments or reading FAQ pages?
- Usage knowledge: What are user and time metric trends on the various platforms you are using, from unique email opens to post likes? What about help desk and support usage?
- Purchasing behaviour knowledge: What makes it into the shopping cart and what causes shoppers to abandon their carts? At what point do they tend to do so? How well do special-offer-after-abandonment efforts work? What results do up-sells and special offers yield?
No, we’re not talking about adult content or nasty rap songs. We’re talking about two different types of data. It’s an important distinction because explicit data yields knowledge that’s immediately actionable, while implicit takes a whole lot of manipulation to turn into knowledge—something best accomplished with tools like machine learning. When you accumulate knowledge, be sure to draw that distinction. Here are the definitions:
Explicit data is that which leaves no ambiguity around customer intent. As examples, someone buys a product, rates a film, or gives a post a thumbs up. It’s clear how they feel about a product, service or idea. There is a direct line from this kind of data to the knowledge that you can act on confidently.
Implicit data leaves you guessing at motives. It’s much more prevalent than explicit data and includes actions like browsing a website, opening an email, viewing a product online without making a purchase, or simply reading an article. Because action does not clearly indicate intent, it requires manipulating before you have the knowledge you can act on.
An interesting study on lead scoring shows that companies with the highest (most effective) lead scores had incorporated at least 3 implicit attributes. The study’s authors suggest that the reason why many companies don’t use implicit attributes is that they are much harder to process/implement—but that machine learning can be used to solve this issue.