BLOG 2 ”Stakeholders”: Data accessibility in AI-driven B2B sales

It can be surprising to realize the number of stakeholders that use, create, and own the data in the company. Stakeholders and their concerns need to be addressed from the start to finish if one aims to utilize data across the company’s silos.

When you want to conduct analysis beyond typical sales or revenue forecasts, such as AI-based analysis, requirements for data sources increases. For example, in the use cases presented in Part I of this blog series (e.g., tailored products), analysis is only possible when one has access to several databases in company, such as CRM, ERP, and IT Service Management (ITSM) systems.

In each of these systems, there are several stakeholders involved—whether they are the data owners, those who create or use the data, or even those who maintain the IT system that holds the data (see Figure 1 and 2).

The current blog (Part II) explores roles of data stakeholders in AI-driven sales. In the blog serie, we have previously discussed about use cases (Part I), whilst upcoming blogs will tackle data wrangling (Part III), confidentiality of AI model (Part IV) and business benefit (Part V). Blogs are published at https://www.innosale.eu/. Please also join our webinar 29.5.2024 14:00-15:30 Finnish time (13:00-14:40 CET), registration link.

Company data siloes

With large companies and corporations, organizational units are often siloed, creating an additional layer of complexity for data access. For instance, sales organizations need to store and utilize customer data, which is typically managed through CRM systems. In production organizations, the focus lies on Manufacturing Execution Systems (MES) and Enterprise Resource Planning (ERP) systems. As long as organizational units do not require data from other units to achieve their own goals and KPIs, cross-organizational data integrations may not be necessary.

The siloed model faces challenges related to data-informed decision-making, the need for product traceability, and Scope 3 emissions reporting.

How do you break down these siloes? In some companies, IT departments are tasked with enabling data sharing across the organization. Another option is to create a small unit involving IT specialists and those competent in data analytics to implement the integration in practice. However, in both these approaches, the main challenge is to motivate organizations and individuals in silos by highlighting the benefits available to them when sharing data. In addition, top management should be able to provide additional resources for the work as well as justification of the business benefit for individual units and the company as a whole.

Data Owner or a Stakeholder?

Why is it important to identify the data owner? In most cases, the data owner is the one who can make a final decision to allow access to the data and can even provide the needed resources to support this action. In principle, the company owns the data, but in practice, it can be said that someone or something “owns” the data. The personnel close to operations should be contacted and address their potential concerns and uncertainties. Stakeholders should find motivation from their own needs to support a data delivery process across silos.

For CRM data, the owner can be a Sales Director or an organization such as Sales or Marketing. The owner for a certain subset of data could be a Sales Manager for a key account customer or a specific market area. Similarly, for ERP, the owner may be the product organization, or the responsibilities could be distributed among many organizations using the system.

The term ‘stakeholder’ is broader than ‘data owner’ and includes other individuals and entities responsible for using, creating, storing, securing, and affected by the data. Nevertheless, any of these stakeholders could become a bottleneck for accessing the data.

Origins of data

Typically, CRM and ERP data are created over the years in daily company operations. Key account managers fill CRM systems with customer information such as name, address, opportunity name, offer contents, price, and sales probability. The Marketing department may prefill contact details of potential customers from market surveys and lead generation activities. Additionally, other company employees than sales persons might enter customer information into the CRM. In some cases, the data can originate from another company, e.g., due to an earlier merger. In this scenario, there can be legal boundaries for using the particular data set coming from merged company databases.

One can also raise a question: Do people who have originally created and uploaded the data to CRM have a say with respect to using the data? The data may contain personal notes or other personal information, which was important at the time but may not be necessary for analysis and could be restricted by GDPR.

In a service or platform business, the data can be owned by a customer’s customer. For example, in phone sales, the customer has a software platform that helps find end-users and their contact details, as well as provides a manuscript for a sales pitch. An example of such a platform provider is the Finnish company LeadDesk. LeadDesk’s platform allows their customers to record and monitor their sales activity’s effectiveness. This customer’s customer data could help LeadDesk improve their platform by adding AI-based advanced analytics functionalities.

Leaddesk aims and perspective

LeadDesk is interested in using AI-driven data analytics to find and optimize factors that contribute to successful sales. By analyzing manuscript and call data, LeadDesk can bring features to their customers that help them identify these factors, allowing them to replicate successful strategies (see Figure 3). This enhanced understanding would help sales leaders improve their sales efficiency. It would also enable LeadDesk’s customers to close more sales with less effort and in a shorter time span.

LeadDesk’s collaboration with VTT on the InnoSale project represents a significant leap toward improving sales dynamics in business. We introduce the project through Customer Success Managers or Sales to potential participants. Once the customer understands the potential benefits of the project and gives their approval, we ask for the specific data needed to make the project a success.

Data security is essential for our customers’ peace of mind. LeadDesk is unwavering in our commitment to data security. We recognize the trust our customers place in us, and we have stringent measures in place to safeguard their data. Here’s how we do it:

• Selective Data Usage: We only use data from selected customers who have agreed to be part of the research project and see potential value in using the insights in their daily operations.
• Anonymization and Internal Security: Personal data undergoes a rigorous process of anonymization and internal securing at LeadDesk before being sent to VTT. We ensure that no personal data is shared without proper safeguards.
• EU Boundaries: All data is strictly housed within the boundaries of the EU, ensuring it remains unpublished and is leveraged solely for AI model training and evaluation.
• Restricted Access: Data access is restricted to those within the InnoSale project team who have pertinent roles, such as data processing for analysis.

Beyond the measures in place for the InnoSale project, LeadDesk’s overarching commitment to data security is evidenced by our adherence to global standards, including our ISO27001 certification.

As LeadDesk and VTT collaborate, our vision remains clear: to continually refine our analytical processes, ensure absolute data security for customers, and work collaboratively with our clients to radically improve sales efficiency through data-informed decision-making. Our journey on the InnoSale project represents just one step in this direction, and we are fervently optimistic about the strides we will make in the future.

Best practices

Here are some examples of how cross-siloed data access challenges were tackled in the InnoSale project:

• Limiting Data Scope: Define a narrow use case or cases to limit the amount of needed data and its sources. Focusing on specific scenarios helps streamline data access.
• Early Stakeholder Identification: Identify stakeholders affecting data accessibility early on. These stakeholders include data owners, business units, and support units. Establish communication with them to address data-related concerns.
• Visualizing AI Benefits: Showcase the benefits of AI analytics to stakeholders. Demonstrate what is achievable with data and how the company or stakeholder can benefit. Leverage expert resources to effectively communicate the advantages.
• Building Trust: Involve cyber security, IT, and legal teams to address trust concerns. Ensuring data security and compliance builds confidence among stakeholders.
• Allocating Resources: Allocate resources from each company silo to support practical tasks such as data extraction, anonymization, and process ownership. Collaboration across units ensures smooth implementation.
• Company Commitment: Highlight the overall company commitment to effective communication with stakeholders. In large corporations, it may not always be clear how data-related activities align with company strategy and revenue improvement.

Authors

Marko Jurvansuu (VTT), Jarno Tenni (LeadDesk), Trang Nguyen (LeadDesk) and Colm Ó Searcóid (LeadDesk)