While ABM can help you target the key decision makers in your accounts, job titles aren’t always what they seem. Columnist Sonjoy Ganguly explains how to factor hierarchy into your ABM strategy.
Account-Based Marketing (ABM) is a powerful way to target decision-makers inside your target accounts. However, your ability to reach the right people depends on your data quality and the tools set to achieve your goals.
But there’s still a wrinkle when it comes to understanding data as it relates to roles inside the account. You see, job titles aren’t always what they seem. That nuance alone can throw off your ABM strategy and approach.
Allowing for rampant title change within your strategy
As today’s companies grow more complex, internal decision-making hierarchies become more specialized. As startups and emerging tech companies compete to retain talent, job titles, hierarchy and org charts reflect the new dynamics of the market. Today’s “Junior Account Executive” is tomorrow’s “Senior Vice President of Account Management and Market Development.”
Combined with non-structured, user-defined titles present in social media, how do you identify your decision-makers? If your prospect has a non-traditional title, it’s very easy to overlook him or her when you’re running a traditional ABM program.
Today’s more intricate corporate hierarchy demands greater levels and degrees of specialty. Thus, an increasingly nuanced view of your target decision-maker’s role is required.
Given the inherent intricacy within today’s ABM initiatives, the latest B2B audience science, which deals with titling in a more granular way, is your most critical asset when it comes to mapping your way into accounts. It will help you break down your decision-maker targets with great specificity, nurture them more thoughtfully, and tailor your content marketing, resulting in more meaningful engagements.
First ask yourself: Who would you like to engage? Is your target audience determined by industry type, region, authority level and so on? What is the spectrum of all their functions? If your campaign is like most, you’ll use a combination of filters to define your audiences.
To find your way to the contacts you care about the most, don’t just think about titles, but develop a taxonomy of decision-makers. Key considerations include industry, specialty and function — with sub-functions being more important than ever before.
Automation and machine learning help
As ABM technology continues to evolve, improved machine learning models to cleanse your data assets can help identify a higher function and also a sub-function, based on job titles, which can assist you with more specific, nuanced targeting. After all, as the diversity and size of buying committees have given rise to the consensus sale, tying your ideal customer profile to specific job titles is the equivalent of searching for a specific hotel room number when you’re traveling.
To hone the methodology behind working this way, it’s important to recognize that each company will have its own unique “title matrix,” the set of titles used within the company. This can be unstructured, or structured into career tracks. But this structure is rarely revealed to outside parties, and it changes over time.
Requirements for mastery
While there are common trends for titles within an industry, and for functions that crosscut industries, ultimately each company decides on its own set of titles. In rare cases, your target account will use a flat titling system. Bell Labs famously assigned all engineers one title, “Member of the Technical Staff.”
So, as a starting point to tackling this challenge, being able to see the title matrix of your account would help you as an ABM marketer to target on authority. However, that is only one of many requirements, including:
- Knowing primary, explicit titles — Individual title conveys information about relative authority. As an ABM marketer, you are interested in influencing key links in the chain leading to a purchasing decision. So getting a lead on authority is key. Machine learning can help level those taxonomies into a more normalized form and grant insight into the most probable chain of authority leading to a decision. With that information, an ABM campaign can precisely target the purchase influencers.
- Understanding sub-functions — The key to beginning to gauge real authority is looking for functional information that makes a person more likely to have a job title, and then match on job titles. Consider working with statistically improbable but frequent phrases. These are rare and quirky but noticeable, and therefore, important.
- Addressing your biases — Further, because you can train a machine learning model only on the data to which you have access, it’s key to get access to well-labeled, statistically representative data sets and maintain these over time. Currently, there is excessive data mostly on technical professions and particular technology companies, and this gives rise to data bias and limits the usability of the data. This bias makes it difficult to make predictions in the areas where you have no access to data. To address bias sources in machine learning, it’s key to identify them and then train your machine against them by developing workarounds.
Next Up: Cracking the code
An intelligent model for titling within an ABM context comes down to a few basic measures to get started:
- Identify your cohort of target account companies. Identifying them means that these companies share a defining characteristic. Even if the target accounts span industries or nations, consider treating these as their own cohort.
- Identify the set of job titles in use for the cohort and the frequency of their occurrence at each firm. This requires access to a large pool of B2B audience data, which your solution provider should have.
- Normalize the cohort titles. Here we identify equivalent titles across companies within an industry, and we distill multiple equivalent titles into one label. The result is a smaller, consistent set of titles.
- Identify hierarchy. Grammatical rules will identify some structure — for example, “senior” is above “junior.” A pool of resumes within the industry will reveal common title transitions as people advance. Also, identifying the frequency of title use within a company (as noted above) will provide a sanity check; executive titles are more rare than junior titles.
A successful analysis will reveal a graph of abstract career paths within an industry, grouped by functional area, with graph edges that indicate probability of seniority. You will also have the real-world titles that map to points on the graph.
Through the magic of machine learning, you have reverse-engineered your target account’s title matrix. Code cracked.
Machine learning is, at its core, just the science of statistics, which is a centuries-old discipline. But machine learning also includes a class of now-feasible techniques for deriving insight from large quantities of data.
by Sonjoy Ganguly
source: MARTECH TODAY