While artificial intelligence (AI) was certainly the top topic of discussion during Forrester’s Technology & Innovation Summit, the conversation appears to have moved on. In preparation for the era of agentic AI, organisations are starting to consider where employees fit; where to use contractors and external service providers; and what tasks should be AI-enabled internally.
In a blog post to tie in with the event, Forrester research director Mark Moccia wrote about how a third of CIOs will adopt “gig worker protocols”, where IT teams comprise AI agents, gig workers and employees with multiple jobs.
In his keynote presentation at the Technology & Innovation Summit, Manuel Geitz, Forrester principal analyst, discussed how business and technology leaders should prepare for this shift. “You start by really understanding which expertise you need to drive your business model,” he told attendees.
For Geitz, IT leaders can get their organisations ready for workflows that may be split between internal staff, external contractors and AI agents by capturing the knowledge using structured data ontologies, to make expertise machine readable. He suggested delegates can then begin to experiment with business models that monetise this expertise on demand, eventually building a platform, where AI agents become the front line for knowledge delivery, supported by humans.
The idea of a gig economy using AI agents is something that appears to be gaining traction among industry commentators. In a recent conversation with Computer Weekly, Jessica Apotheker, managing director and chief marketing officer at Boston Consulting Group (BCG), discussed how the marketing function – which tends to draw on both external and internal expertise – could evolve with an agentic AI workflow. As an example, she discussed the content production workflow.
“There’s a tonne of external people working in the content work group,” she said. “There’s creative agencies, production agencies, localisation agencies. There’s internal people and local marketers, and there’s the tech people. All these people need to come together and reinvent.”
According to Apotheker, this is because AI has the potential to change the content workflow process. IT and business decision-makers need to reconsider what parts of the process they want to own and what parts can be automated, or should be outsourced to a service provider who may well use AI and automation to complete the work: “What is the part of the workflow I think I need to strategically own and transform, and how will that connect with what I actually outsource or potentially automate myself?”
Putting a price on value
Research from BCG suggests organisations that are seeing significant business benefits from deploying AI tend to be AI-first, which means business leaders reconsider the role people have in a business process or workflow, where some aspects can and will be automated with AI.
“Think of an AI-first workflow,” said Apotheker. “You need to rethink what you make and what you buy. It is not obvious that your current make or buy strategy is the one that you need. You just want your contractor to do the automation on their piece of the workflow.”
In a recent podcast, Prem Ananthakrishnan, global software practice lead at Accenture, discussed how the use of AI and agentic AI in business processes is shifting how people think about software.
“There is a fundamental change from understanding that software cannot just be purchased as a tool, to thinking about software as a collaborator that’s driving an outcome for the business,” he said.
This is the next shift in software licensing, one that moves purchases of technological capabilities beyond consumption-based pricing. Mirroring the remarks of the Forrester analysts and BCG’s Apotheker, Ananthakrishnan said: “We still think of buying software as procuring a tool. We need to think about procuring a collaboration vehicle. In my view, IT buyers need to evolve from thinking about procurement to performance and design thinking. Don’t think about buying software anymore. Think about how you’re hiring digital teammates.”
Ananthakrishnan believes these digital teammates will be paid based on outcomes, using what he terms “value-based pricing”.
This is a huge mindset shift, but business and IT leaders can start with something they already have a grasp of: business process outsourcing (BPO) – evaluating which parts of the process are strategic should remain in-house. In the conversations Apotheker has had with organisations that are considering an extreme makeover of their workflows and business processes, she said: “Either you take a BPO approach and fully outsource to somebody, hope that they will transform the process with AI and incentivise on outcomes, or you reshape the process internally.”
For now, Accenture’s Ananthakrishnan noted that token-based pricing and AI credits, which are often applied when purchasing AI-based services, are proxies for value. The more an AI service is used, the more tokens are needed and the more credits are consumed. He said these consumption-based pricing models provide a bridge to leading to an outcome-based pricing model where organisations hire AI agents to take on work.
Ananthakrishnan recommended that IT leaders start implementing business impact metrics, such as linking return on investment to an AI credit model. They might also consider a hybrid model priced on an upfront AI credit, where the supplier is paid a bonus if a certain outcome is achieved.
There is plenty to consider as working practices adapt to include agentic AI – but irrespective of whether AI-enhanced work is achieved internally or via an external service provider, value-based pricing is coming, and people in IT leadership and procurement will need to assess how risk versus return changes when the product or service that is being procured is a probabilistic environment rather than a very deterministic environment.
