3 factors for a successful university-business collaboration
November 12, 2020
By Mark Sierbert, PhD
Personal impressions — and lessons learned — from establishing Elsevier’s DiscoveryLab in Amsterdam
Caption: Dr Georgios Tsatsaronis, VP of Data Science and Research Content Operations at Elsevier, and Dr Paul Groth, Professor of Algorithmic Data Science at the University of Amsterdam, are members of the DiscoveryLab. (Photo by Alison Bert)
A recent European Commission studyon university-business cooperation(opens in new tab/window) suggested that the key benefits in joint research include “innovation with a longer-term horizon, shorter-term problem solving, … data for high quality research and the ability to bring research into practice creating impact.”
Local innovation ecosystems — such as the Amsterdam Data Science(opens in new tab/window) (ADS) network of academic-industry partners — can act as catalysts for regional competitiveness. Campuses can act as “platforms or hubs,” as with the ICAI Labs(opens in new tab/window) for AI in the Netherlands. These in turn drive societal and economic impact for governments.
These hubs are often the result of collaboration between academia and commercial businesses. The benefits are great, but getting such collaborations off the ground can be tricky. So what are the secrets to a great collaboration?
Working with Amsterdam Data Science, Elsevier launched the digital DiscoveryLab(opens in new tab/window) this spring. The lab is the next step in a collaboration that started five years ago, inspired by a joint vision on using AI and knowledge graphs to make research easier(opens in new tab/window). And it’s a testament to what can be accomplished by hard work and cooporation of everyone involved.
In the connected worlds of collaboration and academia, some partnerships are more successful than others. Sometimes these projects can get weighed down by conversations on shifting budget commitments or missing to hire the ideal researcher due to negotiation delays, different legal interpretations or unclear scoping in detail. And in times where speed of innovation matters, people can find they take a long time to get going.
To overcome the challenges, let’s look at the ingredients that make something like this work.
Everybody would agree that trust is important, but the bigger questions are: “Trust in what?” and “What are the trust-building elements that help shape successful collaborations?”
The first thing that comes to mind for me is that personal “click” that gets the partners excited about an idea. But the key part of trust comes when you face challenges and the initial inspiration turns into a detailed negotiation.
Collaboration journeys are “trust building” but also “trust using” for the partners that have to keep believing in the joint intentions and shared ideas so they can find a way out of deadlock situations and see the willingness, support and commitment that will lead to mutually acceptable alternatives. Here, trust manifests itself in being open and honest and respecting each other enough to know what is acceptable and helpful.
There’s also trust in the agreement itself – how much the partners need to fix contractually, what expectations to set into delivery capabilities or which conditions to ask for up-front in times of conflict.
In this realm, there are no black and white answers. Yet there are ways to enable trust. It is in the nature of innovation and research partnerships that not all eventualities can be pre-defined or clarified. And the more regulation is put up beforehand, the less flexible the partnership will be when it needs to explore, adapt and adjust in order to strive for excellence and mutual benefit.
As it can’t be a “free cheque” either, that’s why partnerships and trust grow over time. It helps partners understand the different ways of working and institutional circumstances and to establish a regular dialogue to address and clarify issues. Working along those collaboration principles, next to the contractual base, provides the flexibility and energy to excel together. Literature consolidates this into the Trust Equation(opens in new tab/window): Trust = (credibility + reliability + intimacy) / self-orientation.
Looking back at our terrific experience with Amsterdam Data Science, I would say three success factors played a role:
1. Build credibility, reliability and shared excitement as you get to know each other’s unique abilities.
Getting to know each other’s capabilities in different formats over time creates a unique experience that helps clarify which partners are well suited to which projects. For these partnerships, where ideas surface based on previous collaborations, there are typically no big investments available, so we use joint funding applications, master’s student projects, and smaller contract research work, along with managers interested in experimenting. This takes time and needs a farther horizon, but it builds a strong foundation.
With ADS, we helped shape an ecosystem around this work that allows all partners to explore ideas with each other and share the joint excitement before making bigger investments. With the DiscoveryLab, we consolidated and connected various joint projects and found excitement around research knowledge graphs.
2. Differentiate business/research priorities and legal requirements to minimize “self-orientation.”
The most frequent complaint I hear across collaboration set-ups is about legal delaying and complicating things. Involving legal too early in the process, as people often suggest, can actually complicate the partnership at the wrong time, when business questions are not worked out. If they’re involved too late, it can delay the process at the end, as key aspects might have been overlooked. So how do you find the right balance, and how do you actually benefit from the legal queries raised in building trust? What is the right level of detail to specify in the contract? What type of governance should be installed to handle details? What are cultural and functional differences or just fair differences of interest?
In essence, few of the “legal” claims are unnecessary. In my experience, what holds the process back is that business and academic priorities are not detailed and clear enough to state what is key to the business/research and what is optional. For example, with IP, patents and know-how, what risk is the business willing to take, and will the business be able to absorb the research, turning it into products and platforms? There are also questions surrounding the level of commitment the university and researchers can make in terms of liabilities or AI ethics. And there are differences in style, such as the level of detail that needs to be agreed upon and what can be left to later negotiations.
So it’s the question of how you formalize working with uncertainty and whether you’re open to defining the approach, process, principles of work and resources rather than the outcomes. Existing agreed contract work, like the pre-agreed ADS research contract, helps to build upon and align the work of all the partners — which in our case were the University of Amsterdam, Vrije University, Elsevier and the government. Everyone has their own legal templates, but it can be a challenge if we lack guidance to match the business/research needs and priorities to properly fill those. Naturally the less clear outcomes and processes are, the higher the appearance of legal issues — even if they are not of legal nature.
And yes, reminding ourselves regularly of our shared excitement and putting ourselves in each other’s shoes to reduce “self-orientation” always helps in getting us back on track.
Let’s shape this shared knowledge base for the ADS ecosystem with an experienced team of people from the organization who are experts in AI/data science collaborations, including lawyers, IP experts, data sharing experts and AI experts.
3. Use facilitated communication to connect the interests and expectations that must converge to build trust.
Communication is yet another over-stretched success factor that will unfold differently each time. We’re all used to dealing with different communication styles, but what’s relatively recent is the external dynamics, especially in AI, with continually changing expectations, regulations and set-ups to align with.
To progress in good spirits, it was helpful to have people in the partnering group to facilitate, drive and support the dialogue and negotiations.
In our case, subsidy requirements kept on adding to the complexity, with AI ethics principles being added, reorganizations happening, alignments getting revised and budget commitments questioned — probably normal challenges in such a process. We tried quicker and bolder decision-making to speed up the process and reduce the likelihood of new events interfering. And ultimately we learned that a good foundation pays off afterwards.
ADS is recognized for its network and ecosystem culture. Elsevier is an established network partner of the scientific community. In the same way editors host and facilitate a research group or ADS facilitates a data science network, we made use of partner managers (like me) to take care of the process and help unwind difficult situations. Sometimes milestones, such as the launch of the Netherland’s national AI strategy, and sometimes personal talks to provide transparency helped us to adjust and move on. Yes, ideally there would be more transparency of government decision criteria, formal requirements or budget preferences. Every time I tried to pre-define those, I learned that it’s good to strive for, but it’s better to be resilient and deal with it in a partnering spirit.
There is nothing better in a partnership than winning together.
Now that we’ve entered a new phase, we’re excited to experiment with new ways of working – sharing desks on campus and at Elsevier (at least after Covid), exploring ideas in joint reading clubs and cracking the next research challenge. We already see more partners joining our teams, and we are happy to continue the conversation.
Related stories and resources
(Elsevier Research Intelligence)