Artificial Intelligence is expected to impact many different aspects of marketing in business to business (B2B) organisations, from advertising, to product development and customer service, to name a few. Though, for those benefits to materialise, firms need to have technological and technical capabilities in place by either developing them in-house or using external suppliers. How B2B organisations approach the deployment of AI in marketing will shape their position vis a vis other stakeholders in. the market and, with it, their competitive advantage.
Brendan Keegan, Dorothy Yen and I investigated the drivers and forms of adoption of AI marketing applications in the B2B sector, and their impact on the relationships within the network, through the lens of power dependence theory.
Specifically, we investigated the following research question “How does the adoption of AI marketing solutions affect the power dynamic between focal firms, AI suppliers (small to medium-sized), AI tech giants (e.g., Google Cloud, Amazon Web Services), and customers within the service network?”
We investigated this question via in-depth interviews with 8 B2B firms who are deploying AI solutions, 5 AI suppliers, and 7 AI advisors, who each had between 3 and 25 years of experience in the AI arena. The interviews explored adoption motivations, experiences of deploying and using AI B2B solutions, the impact of AI adoption on network relationships, and future development plans. One of the interviewees described the situation as a dance, which resonated with our own analysis of the interviews. Hence, we adopted the metaphor of the AI B2B solutions market as a tango dancefloor, where the various players in this market are the dancers.
We broke our analysis into four stages of the adoption process:
- Entering the dancefloor: The decision to adopt AI
We found that adoption is driven by three factors. The first two had been mentioned in the literature, already, and are: the perception that AI can save costs and lead to new insights. The third one is (perceived) pressure by customers, leading firms to embrace AI in an attempt to be seen as a competitive and credible business partner. By and large, firms lack the in-house expertise to develop – or even take advantage of – AI solutions.
There are two main types of suppliers in the AI solutions markets: small ones with expertise in a specific application, and large ones with largely generic solutions. The former usually lack the brand recognition and computing power of the latter (e.g., Google).
2. Learning the steps: The first experiences with AI
Because of the lack of in-house expertise, firms tend to outsource their AI solutions. Among our interviewees there seemed to be a preference for large suppliers because of brand recognition, even when they suspected that niche suppliers might provide a better technical supplier. They enter into contracts that, typically, require generalised transfer of customer data (as opposed to carefully selected and justified batches of data).
To gain recognition in the market place, small suppliers focus on gaining recognition for highly specialised, proprietary algorithms. Another strategy adopted is to partner with their larger counterparts to gain access to their expertise as well as to benefit from the halo effect of being associated with those brands.
3. Negotiating the dancefloor: Ongoing use of AI
Firms report benefits from adopting AI, but mainly in very specific applications such as targeted advertising, where there is abundance of data, the task is repetitive, and the downside of false of positives is relatively small. Success is more limited in product innovation, and rare in customer service or applications requiring justification of how a certain decision was reached. At the same time, there is a growing unease about the extent of data sharing with, and the growing dependency on, AI suppliers.
The accumulation of data (via contracts with their client firms) helps suppliers grow exponentially, as they are able to strengthen their products (e.g., algorithms), and strengthen their reputation vis a vis large suppliers. However, there are also some signs of “scope creep”, with suppliers positioning their AI solutions as suitable for more and more types of tasks. There are also cases of suppliers taking advantage of clients’ dependency on their solutions to renegotiate contracts.
4. Future Steps, Beats and rhythms: Future directions
Having dipped their toes in the field of AI solutions, and gained a sense for what they can bring to the organisation, some firms are starting to develop their own in-house AI teams. Other firms are likely to continue relying on external suppliers, though through a more carefully curated network of suppliers.
Small suppliers that are unable to secure long term contracts, or that fail to develop a distinctive reputation in this marketplace, are likely to be squeezed out of the market. They may end up being acquired by another, larger AI supplier. Access to data and contracts with small suppliers give the large companies even more power in various forms: economic, technological, technical, and legitimacy.
Some move towards protecting customers’ rights and their data, via GDPR and other tools. Though legitimacy of data ownership is yet to be defined and claimed among data creators (customers) data collectors (focal firms), data processers (AI small suppliers) and data storage platforms (tech giants).
This work has some implications for theory, such as identifying a new driver of AI adoption (i.e., reputation), and the exponential effect of data as a source of power. In addition, this work has implications for marketing practice, which we encapsulated in the following managerial recommendations.
Implications for B2B marketing managers
When calculating the costs of the AI solution, managers need to account for additional costs such as equipment, processing power and skills development.
They also need to account for the reputational cost of delaying AI adoption vs the reputational cots of potential technology failures (false negatives). Cheaper solutions tend to be less customised for the specific firm. Small suppliers tend to offer more creative solutions than tech giants.
To pre-empt suppliers’ attempts to take advantage of firm’s dependency and increase their fees, contractual agreements should contain performance targets and lock-in rates. Firms also need to consider how their proprietary data sets are going to be used by suppliers in the future and/or over and above their current contract.
Implications for niche AI suppliers
It’s the ability to create customised solutions gives small players a competitive edge. So, these suppliers will benefit from engaging in value co-creation with their clients.
Niche suppliers need to assess whether / how entering collaborations with tech giants can provide them access to large databases and facilitating platforms.
Implications for large AI suppliers
There is no denying that these players are in an advantageous position by virtue of first mover advantage and or their superior processing power.
This firms can materialise the promise of exponential growth enabled by the continuous access to data, though supporting AI start-ups via platform collaborations and model-sharing. This will secure their network leader positioning, and facilitate their continued dominance in the AI marketplace.
The paper is available here. As usual, I am looking forward to hear what you think of this paper, and delighted to discuss how I can support your work.