You are an early stage technology start-up and want to find out who are your ideal customers!
Having been part of a small technology company, one of the biggest challenges we faced was to identify the right customers for our B2B business model. We are building a platform with embed cutting edge technology that is entering a new market and as such there is no data about who would be the ideal customers.
What is the best segmentation strategy?
We have taken a traditional approach to segment the market and we looked at firmographics segmentation: industry, location, size, status or structure, and performance. We looked at the behaviour consumption patterns and extracted the firmographics data to identify the key drivers for the sales funnel. The real customers have not been the most intuitive ones. What data proved to be most relevant?
Identifying the ideal customers to drive hyper-personalized content, ROI and sales
Why are the best performing accounts the best performing accounts? The right segmentation can help a company identify the right customers and this in turn will drive ROI and increased sales.
What is the essential data that can help marketers identify the right accounts? The firmographics is a good basis or starting point to identify this data. In addition, a propensity model could assess the potential identified ideal customers resulted from the firmographics analysis. This model can assign individual probabilities and predict whether a company might buy a product. It will look at the past behaviours of the companies in relation to a certain action in order to assign individual probabilities that they will perform that action. “Probabilities are assigned by estimating each covariate’s (industry, location, etc.) contribution to an actual purchase. “(Bally Kehal, 2019)
“Irrelevant data will make you waste your time” (Bally Kehal, 2019) To become even more relevant, web-scrapping services can help tailor the products and the content marketing even more to go beyond the company level, to the buyers’ level. “B2B buying decisions often rely on a team of decision-makers.” (Bally Kehal, 2019)
Content playlist is a second technique that can drive hyper-personalized content marketing. It providex all the content about the product upfront, so that a company can learn more about their customers: problems, needs or solution interests.
The personalized content will drive precise analytical insights and personalized content and ultimately it will drive better conversions.
“Informative and relevant information accelerates the B2B buyer’s journey, just as it does for the B2C journey.” (Bally Kehal, 2019)
Segmentation and the data variables to consider
We have mentioned above the importance of the market segmentation to identify the ideal customers and the different techniques to deliver more relevant and targeted communication that is relevant to the segments. This will form the basis of an effective growth hacking strategy which has at its core customer centric marketing strategy.
But let’s speak about the data input necessary to identify the right segmentation. “Data is the new oil!” (Jan Teichmann, 2019). Now, vast amount of data can now be processed with the aid of machine learning algorithms (ML) that can scale and automate this data. For B2C models behavioural data such as click-streams, event and search data from e-commerce website or apps can unlock innovative use cases. (Jan Teichmann, 2019) However for B2B models which is the behavioural data that is the most relevant?
The different geographic, demographic and psychographic attributes of your customers don’t apply in a B2B model. Instead data such as industry, location, size, status or structure, and performance (firmographics that have an impact on behaviour) is more relevant and can be extracted from the targeted businesses behavioural data with ML algorithms in order to offer unique insights.
In addition to the firmographics (industry, size, location, etc) other factors have to be taken into consideration for B2B marketing segmentation: operation variables (technology, user non-user status, customer capabilities) purchasing approaches of the companies (, situational factors and personal characteristic. This data can help develop a precise data model. This model can consider operating variables such as what customer technology to focus on, which users to focus on (heavy, medium, light and non-users) and customer capabilities such as demand intensity of services. Purchasing approach factors have also been taken into consideration to ensure we are focusing on the right customers: are the companies decentralised on centralised purchasing organisations? Other questions are if these companies are engineering of financing dominated? Should we serve the companies that we want to work with or the ones we have a strong relationship with? What type of policies do these companies have in place? Should we serve companies that are looking after quality service or price? Situational factors also play a major role in defining the input data and add further impact factors details such as the urgency of products we provide, the applications of our products and the desirable size order that we prefer to focus on.
And lastly, is it important to serve companies that have the same values as us, are loyal to their partners and what is their risk aversion? All the above parameters have been taken into consideration when building the data model to drive the market segmentation. (Philip Kotler, 2019)
The needs based segmentation is at the core of our data segmentation model: needs, segment identification based on needs, segment attractiveness, segment profitability, segment positioning, segment “acid test” and marketing-mix strategy. (Philip Kotler, 2019)
The use of data and ML algorithms to process it has the advantages such as having relevance of customer attributes and being transferable to other type of data classification algorithms. It has the disadvantage that is not directly human interpretable. (Jan Teichmann, 2019).
Bibliography
Bally Kehal (2019) ‘Three Approaches to B2B Segmentation: Firmographics, Propensity, and Role-Based Marketing’, Three Approaches to B2B Segmentation: Firmographics, Propensity, and Role-Based Marketing, 13 September. Available at: https://medium.com/aiautomation/three-approaches-to-b2b-segmentation-firmographics-propensity-and-role-based-marketing-34ab325c567e (Accessed: 19 July 2020).
Jan Teichmann (2019) ‘New gold standard: using machine learning to derive a user and product segmentation from behavioural data for a marketing STP strategy’. Available at: https://towardsdatascience.com.
Philip Kotler (2019) Marketing Management. Pearson.
by Anne Spulber
