"Unleashing the Power of Partnerships in the AI Ecosystem: The Role of Business Development"

I have had a lot of conversations recently about Partnership and Business Development teams in various gen AI and traditional AI/ML companies. Often times it is well understood that partnerships are very necessary to drive product differentiation, adoption and ecosystem growth. However the stakeholders have unclear notions of exactly what they need the partnerships team to do. Here are some thoughts and observations on partner types, differing motivations and desired outcomes. Aligned with this is the key roles various teams, i.e., Product Managers, GTM teams, Sales, Marketing etc. have in delivering a successful partnership.

First to ground the discussion, lets discuss the gen AI stack as a way to understand the complexity we are dealing with.

Gen AI Ecosystem and Modern AI Stack

The Gen AI ecosystem, characterized by its multi-layered complexity, spans from foundational IT infrastructure providers, data management/Vector DataBases , Orchestration and Management layers and finally the application layers that manage Observability, Model governance etc. A lot of these are new frameworks and paradigms that have emerged fairly recently (Transformers et al) and seek ecosystem growth to establish themselves as de factor frameworks in the fast emerging gen AI lifecycle.

At the lowest level the compute and foundational layers the primary focus is on Private and Public Cloud infrastructure. These provide Compute, Storage and Neworking. On the hardware level Nvidia GPUs, CPUs companies like Intel, AMD and AI accelerator startups provide innovative architectures to drive compute at scale. There are capacity contraints (production and availability) in the industry today especially with GPUs, Memory chips. Partnerships take an interesting role in this domain, with large hyperscalers like AWS, GCP, Microsoft Azure using partnerships to ensure significant allocations for their cloud infrastructure.

Foundational LLM providers like Open AI, Mistral, Meta, Anthropic and Google have created a number of models with millions in R&D. Given the compute intensive nature of develop foundational models, these companies are closely aligned with Google, Microsoft (Azure) and AWS. These providers now seek to monetize the LLMs either through commercial agreements with application developers, enterprise software developers like Oracle, SAP etc. or the end consumer with availability through API end points.

The next layer, is the data layer that includes traditional data lake companies like DataBricks and SnowFlake. Vector DataBases and Embedding Model providers like Pinecone, Cohere, Redis, that manage data access, enable building data pipelines are a crucial new segment in the market (arguably vector DBs have existed for a while). Vector DBs and Embedding Model Providers are the de facto part of the LLM lifecycle, however there is increasing focus on partnerships between standalone companies and traditional DataBase companies — a key requirement driving this is the need to co-host contextual data and queries within the vector space so frameworks like RAG can be used seamlessly.

The model deployment and orchestration layer includes companies like LangChain, HuggingFace. Especially as AI Agent workflows are becoming more common place, there is a need for strong governance and tools to allow LLMs, tools, scripts and traditional IT workflows to interact and create intelligent workflows. Finally the highest level is for application layers that manage Observability, Model governance etc.

At this point, it is worth mentioning that this view is not static but incredibly dynamic. Most gen AI companies are looking to further integrate with creation of complementary technologies. For e.g., Langchain is now a commercial entity so they are moving beyond just orchestration framework to provide AI agent governance tools etc. Cohere has made significant strides beyond VectorDB of choice.

Partnership and Business Development archetypes

As we look at the modern AI stack above, the primary focus is on technology partnerships where 2 or more partners can bring together complementary technologies and deliver a more complete or a fully complete solution that drives business value to the end customer. For example Pinecone, one of the industry leaders in Vector Databases, a key technology is actively looking to partner with various Multi-Cloud and Data Warehouse vendors to create a better value proposition for end customers. However to build a product and deliver a business outcomes for end customers, there are a lot more partnerships that are needed.

These fall into 3 main types: “Build Together”, “Co-Marketing/Co-Sell Together”, “Drive Consumption Together”. For the purpose of this discussion I am leaving out the Resellers, VARs, Distributors ecosystem (the so-called “channel”) out of this article. That is a very complex system with intensive margin pressures.

A. “Build Together” Model

The “Build Together” model in the Gen AI ecosystem epitomizes the essence of strategic collaboration, where companies unite to leverage their respective strengths, create innovative solutions, and construct integrated workflows. This model is particularly effective in developing vertical solutions tailored to specific industry needs.

Industry Examples

• OpenAI and Microsoft Partnership: OpenAI’s collaboration with Microsoft exemplifies a powerful synergy where Microsoft’s Azure AI supercomputing infrastructure supports OpenAI’s ambitious research goals. This partnership not only accelerates AI development but also enhances Microsoft’s position in the AI market.

• NVIDIA and Healthcare Partnerships — NVIDIA’s foray into healthcare through partnerships with medical companies exemplifies the “Build Together” model. By combining NVIDIA’s GPU technology and deep learning expertise with healthcare companies’ clinical knowledge, they have developed advanced AI-driven imaging and diagnostic tools. These collaborations have led to breakthroughs in medical imaging analysis, improving the accuracy and speed of diagnoses.

• Cohere, Langchain and Industry Collaborations — Langchain, a platform specializing in language model applications, has engaged in various partnerships to integrate its AI solutions across different sectors. By collaborating with companies in finance, legal, tech, and more, Langchain has been instrumental in developing tailored AI applications that automate and optimize language-based tasks, from document summarization to customer service enhancements. These partnerships are quintessential examples of how AI can be vertically applied to solve domain-specific challenges.

• Hugging Face and Broader Collaborations — Hugging Face, a leader in the open-source AI community, has established numerous partnerships across the tech industry to foster the development and application of machine learning models. Collaborating with companies like Amazon Web Services (AWS) and Google Cloud, Hugging Face has made its state-of-the-art models more accessible and scalable.

Roles and Team interactions

In the “Build Together” model of partnership, the Product Manager crafts the partnership’s strategic framework and spearheads negotiations to ensure technical and project alignment. He/she also leads joint product roadmap discussions with a view to collaborative innovation. Often there is engineering trade-offs, R&D cost discussions and mutual business case development if significant feature changes are in scope. Of course there are different product manager archetypes — a purely technical PM vs a end to end owner in which case the PMs role is far more extensive.

The Go-To-Market (GTM) Manager evaluates market fit, pricing, and aligns sales strategies across both organizations, ensuring cohesive market entry and customer engagement. Marketing Managers then develop a compelling joint value proposition and create materials that effectively communicate the benefits of the partnership to the target audience.

Supporting these core functions, Corporate Strategy ensures alignment with broader business goals, while Training & Enablement staff ensure all field personnel are enabled on the partnership value proposition. Financial analysts complete the picture by assessing the economic impact of the collaboration, ensuring the partnership’s financial health and sustainability. Together, these roles integrate to enhance product value, drive market adoption, and contribute to ecosystem growth.

B. “Marketing/Sell Together” Model

The “Market & Sell Together” model in the Gen AI ecosystem involves strategic collaboration between companies to amplify their marketing and sales efforts. This model encompasses various facets like pricing and packaging, co-selling models, and joint go-to-market (GTM) strategies, each tailored to harness the collective strengths of partnering entities.

• Pricing and Packaging — Companies must align their pricing and packaging strategies to ensure offerings are attractive and competitive. This often involves creating bundled solutions that combine products or services from both partners, offering better value propositions to customers. This requires harmonized pricing strategies that reflect the combined offering’s enhanced value.

• Co-Sell Models — Often the pricebooks for each partner detail the pricing structure of products and services. Partnerships may involve cross-listing solutions in each other’s pricebooks, ensuring sales teams have clear guidelines on pricing combined solutions. This cross-listing facilitates transparency and simplifies the sales process when offering integrated solutions to customers. Different co-sell models can be employed, ranging from referral agreements to revenue-sharing models, each with its challenges like revenue attribution, sales credit, and channel conflict management.

• Joint GTM Design, Marketplaces — A joint GTM strategy is essential for unified market approach, necessitating shared objectives, target market definitions, and agreed-upon sales tactics. Marketplaces, like the AWS Marketplace or Salesforce AppExchange, provide platforms for companies to co-list their solutions, enabling customers to access and deploy integrated offerings easily.

Industry Examples

• Google Cloud and Cohere — Google formed a partnership with Cohere, a generative AI company founded by former Google Brain researchers. Cohere, which develops natural language processing (NLP) models, utilizes Google Cloud’s infrastructure to deliver its AI solutions. This arrangement allows Cohere to leverage Google’s robust cloud services for deployment while participating in joint marketing efforts to promote their NLP products.

• Hugging Face and AWS — Hugging Face, known for its vast repository of open-source AI models, collaborates with Amazon Web Services (AWS) to promote and sell its solutions. This partnership benefits Hugging Face by utilizing AWS’s cloud infrastructure and marketplace, offering Hugging Face’s models as part of AWS’s machine learning offerings. AWS supports Hugging Face with marketing strategies and global sales efforts.

• IBM Watson and Box — IBM Watson and Box have teamed up to enhance content management with AI. Box integrates IBM’s Watson AI technology to offer clients advanced content analytics, enhancing the capabilities of Box’s cloud content management solutions. This partnership includes joint marketing and sales strategies, where both companies benefit from shared branding and co-selling to each other’s customer bases.

Roles and Team interactions

In the “Market/Sell Together” model of partnership, the Go-To-Market (GTM) Manager and Marketing Manager evaluates market fit, pricing, and aligns sales strategies across both organizations. Marketing Managers then develop a compelling joint value proposition and create materials that effectively communicate the benefits of the partnership to the target audience.

Sales Operations or Sales Ops takes a stronger role in defining how the two partners will align during the sales cycle (typically defined in terms of SFDC stages — awareness, customer evaluation, sales initiation, sales closure and customer hand-off). They also ensure the various field teams have aligned incentives, activities and regular discussions to ensure partners are working well together. These are typically in the form of regular touchpoints, QBRs (quarterly business reviews) etc. Financial analysts complete the picture by tracking sales targets, quota setting (in collaboration with Sales Ops), forecasting etc.

Sales Training & Enablement staff ensure all field personnel are enabled on the partnership value proposition, the sales cycle expectations, R&Rs, and have at their disposal the latest methodology that is expected to be followed (typically this has been defined by the sales ops teams).

In conclusion, the “Market & Sell Together” model demands a nuanced approach to pricing, packaging, and selling, with a strong emphasis on collaboration and strategic alignment to overcome challenges and capitalize on the combined market strengths.

C: “Drive Consumption Together” Model

According to IDC, the global spending on generative AI (GenAI) solutions is projected to reach $143 billion by 2027. Gartner predicts that by 2026, over 80% of enterprises will have utilized generative AI applications. The primary challenge being addressed here is the lack of industry capability in building and operationalizing LLMs, shortage of trained individuals across these new gen AI technologies like Langchain, Vector DBs and building end to end applications. To address this demand, there is a need for human capital — either in-house or through partners.

In the “Drive Consumption Together” model, Global System Integrators (GSIs) and Managed Service Providers (MSPs) play a pivotal role, collaborating with generative AI firms to enhance service offerings and drive usage. Global System Integrators (GSIs) are large organizations with tens of thousands of employees globally, such as Accenture, Deloitte, and IBM. Collectively they can mobilize incredibly fast to address this demand.

Industry Examples

• Accenture and Google Cloud: Accenture, as a GSI, partners with Google Cloud to integrate generative AI into business solutions, driving consumption across various industries by offering tailored AI services that improve operational efficiency and innovation.

• Deloitte and AWS: Deloitte collaborates with Amazon Web Services (AWS) to deploy AI-driven analytics solutions, leveraging AWS’s machine learning capabilities to drive consumption in large enterprises, focusing on sectors like finance, healthcare, and retail.

• IBM Managed Services and Watson AI: IBM’s managed services integrate Watson AI capabilities to enhance their service offerings, driving consumption of AI solutions in enterprises by providing managed AI services that are scalable and industry-specific.

Roles and Team interactions

In the “Drive Consumption Together” model, Product Managers, GTM teams, and other key players collaborate closely to maximize product adoption and utilization. Product Managers define and refine the AI service offerings, ensuring alignment with market demands and partner capabilities.

GTM teams develop and execute strategies to market and sell these services, creating materials and campaigns that resonate with target audiences and align with sales efforts.

Meanwhile, roles like Corporate Strategists, Technical Integration Specialists, and Customer Success Managers provide overarching guidance, ensure seamless technology integration, and foster customer engagement and satisfaction, collectively driving the consumption and success of the partnered AI solutions.

Organizing a Partnerships and Biz Dev team

Organizing a Partnerships and Business Development team in the generative AI industry requires a cohesive strategy that aligns product offerings, sales, and marketing efforts towards common goals. Product alignment ensures that partnerships are in sync with the product’s roadmap, leveraging AI capabilities to meet market needs. Sales and marketing alignment involves coordinating strategies to promote and sell the AI solutions effectively, ensuring unified messaging and market approach. Key to this organization is defining clear measures of success, such as revenue impact, market expansion, and partnership vitality, to continuously evaluate and steer the partnership activities. This integrated approach enables the team to navigate the dynamic AI landscape, fostering partnerships that drive innovation and growth.

Other considerations

In addition to the core functions of a Partnerships and Business Development team, other considerations like the choice of systems, tools, methodologies, and governance play a crucial role in the gen AI industry. Effective systems and tools, such as CRM platforms and AI market analysis tools, are essential for managing partnerships, tracking progress, and analyzing market trends to inform strategic decisions. Establishing clear governance frameworks helps in maintaining transparency, managing risks, and aligning partnership objectives with the company’s ethical standards and regulatory requirements, ultimately sustaining long-term success and trust in the AI ecosystem.