Is Transparency Becoming Non-Negotiable?
I recently spoke to Tom Lawrence, founder of MVPR, a team of senior PR practitioners with an agentic platform they've built in-house, called the PR Operating System, supporting execution. They're both a services and a product company, in that they deliver a full PR service to some clients while others license the PR Operating System.
What does the PR Operating System do, and how did it come about?
The platform provides a space for transparent collaboration and execution of all activities you’d traditionally expect to get from an in-house PR team or external PR agency, like tracking relevant journalists and managing email exchanges, using data to develop PR strategies and content, or speeding up execution of proactive and reactive PR work.
We started by looking for basic efficiencies, like how to turn a 30 minute task into a 15 minute one. You don’t necessarily need AI for that, it can often be done via a simple workflow automation. What you do need is a thorough understanding of how things are done currently.
We put a lot of effort into analysing our processes, which enabled us to spot opportunities for improvement. At the same time we realised it was vital to develop a clean data architecture.
Can you give us an idea of what that looks like?
Each major capability in our product now operates through a dedicated microservice which exposes APIs designed to excel at a singular function. This could be finding relevant journalists, analysing campaign objectives, or orchestrating email delivery.
What makes this architecture powerful is the protocol we’ve built on top of it. Our system of AI agents uses this to decide which API to call based on the workflow it’s been assigned. This gives our operating system highly granular control, with each step in an agentic workflow requesting exactly the data it needs from the right microservice.
Designing this kind of system isn’t easy, but now that it’s in place it provides huge flexibility and means we develop new products and features faster, because each part of the system is modular and reusable. It also enables full transparency with our clients, which is a huge unlock. For example, having clean data enables us to attribute all the way down the funnel.
Was selling the platform as a product part of the plan from Day One?
As users ourselves, we immediately know if a new feature saves us time. We always knew if we built something that works for us, there would be potential to offer it to clients.
How does owning your own platform impact your business model?
When we started MVPR we used to sell clients a set number of hours per month. Using the PR Operating System, we found we were meeting the client’s brief in a third of the allotted time.
We’re fully transparent about how services are delivered. Early on, a client challenged us that we hadn’t used their full allocation of hours, even though we’d fully delivered on the brief. On that occasion we gave the client a refund. Then we thought, let’s move to output based billing.
How is adoption progressing?
As with any new way of working, you need to motivate people to adopt it. PR people are problem solvers, we're great at coming up with solutions under pressure. We’ve found the most effective approach is to impose time constraints, then provide the tools that enable people to adapt.
In-house teams love working with the platform, however agencies have been slower to adopt. MVPR’s business model is different to the hourly rate model, and we’ve found that agencies can struggle to adapt their model.
Educating the market is an ongoing challenge. Most people think AI-assisted PR is a next year thing but if you can do it today, why wouldn’t you use this?
Trust in AI is in flux right now. How do you contend with that?
No one will use an agentic system they don’t trust.
User interactions are what builds trust. Using the AI should feel like working with a teammate.
The ability to deliver this experience comes down to the quality of the discovery work and the constraints built into the system. The team designing it needs to know “At this point in the workflow, client approval is needed” or “More context is needed here”.
What advice can you offer other professional services firms looking to augment human workflows with AI?
The best applications of AI are bottom-up. You need people who are close enough to the process to know it in intricate detail. However, they don’t necessarily have the experience to articulate the real problem or opportunity.
So you start with the leaders in a business and go deeper with the why - why do you want this?
The data transformation part is critical. It’s painstaking work, but if you don’t have it, applying AI to your data won’t work. Most companies probably need to bring in external expertise to get this done.
Any final thoughts?
We’re four years into building MVPR and two years into the journey of implementing AI. It’s a bit like driving into the fog. You know what’s behind you, and you can see what’s immediately ahead, but not further into the distance.
The biggest problem for most of the incumbents in our industry is that they didn’t get started two years ago. Now the advantages of applying AI to clean data are compounding. The gap between us and incumbents is significantly larger than it was two years ago.
What I’ve Learnt About (B2B) Trust
To answer this, I looked back over some projects I supported B2B clients on over the past few years.
As I went through this exercise (summarised below), it became clear that trust was often at the heart of the client’s challenge.
Retailer
Objective: Deliver a consistent customer experience at partner-owned sites.
Challenge: Little power to dictate standards due to fear of partners defecting.
Solution: Invest in education and incentives for partners and staff.
Parts manufacturer
Objective: Better (digital) management of the distributor sales channel.
Challenge: Distributors were suspicious, seeing digitisation as an attempt to gain visibility of their customers.
Solution: Create a mutual win by enabling distributors and outlets to easily track orders, earn points and redeem rewards.
Equipment manufacturer
Objective: Capitalise on explosive growth in data centres.
Challenge: The data centre ecosystem saw them as a telecoms specialist.
Solution: Thought leadership to ‘own’ an emerging trend at the intersection of both sectors.
Logistics company 1
Objective: Drive shipping volume without diluting brand positioning.
Challenge: E-commerce platforms control the customer experience, and were unwilling to share data.
Solution: Thought leadership targeting enterprise-grade platforms, showing how tight integration enables a premium customer experience.
Logistics company 2
Objective: Shift from selling services to selling subscriptions to a new SaaS product.
Challenge: Customers were unaware of the product and sceptical about AI-enabled inventory planning.
Solution: Build awareness and credibility by putting the company’s domain experts in the spotlight.
Logistics company 3
Objective: Offer specialised services and move up the value chain.
Challenge: Building trust with customers in regulated sectors is painstaking and slow.
Solution: Use existing tools and warehouse equipment to enable customers to verify process adherence without the need for video surveillance of staff.
Standards body
Objective: Drive uptake of industry standards for materials provenance and sustainability reporting.
Challenge: Suppliers were wary of trade secrets leaking.
Solution: Design a proof-of-concept enabling industry participants to share facts and proofs rather than raw data using open technology standards.
Trade association
Objective: Increase adoption of new technology standards.
Challenge: Scepticism and misconceptions surrounding the underlying technologies.
Solution: Focus on myth-busting, analyst relations and highlighting the value for specific industries.
IT services provider
Objective: Establish a compelling digital presence and drive lead generation.
Challenge: New company with low market awareness.
Solution: Position the CEO and communicate their expertise through consistent, visible thought leadership.
Looking down this list of projects some recurring themes and trust-building patterns jump out:
Lack of trust is the single biggest barrier to digital transformation
Education and incentives are needed to drive adoption of digital solutions
This starts with engaging internal and external stakeholders, and making the business case
Open technologies including verifiable credentials and Privacy Enhancing Technologies (PETs) have an important role to play in accelerating supply chain digitisation
Creative use of existing tools can also enhance trust
Some of these projects identified technical requirements such as identity verification or sharing cryptographic proof of origin or compliance, to enable process automation or boost trust in existing processes.
Nevertheless, success ultimately hinged on building human trust - whether with staff, customers, partners or the wider industry.
My key takeaway is that trust building is a multi-faceted, ongoing endeavour encompassing stakeholder engagement, technology enablement, aligned incentives and visible leadership.
The payoff is faster, more successful digital transformation.