Reinventing Media Intelligence: AI, Innovation and Integration

The rise of AI is transforming the media intelligence sector but as Newton's Phil Lynch and David Barrowcliff explain, there have been technological shifts before. The winners will be those companies who embrace both the potential of new tech and the lessons of the past.

The media intelligence sector’s diary is getting rather full. Social listening platforms Brandwatch and Crimson Hexagon have completed their merger, Ipsos is buying Synthesio and WPP is selling off a majority stake in its media insights arm Kantar. Clearly something is going on. 

In large part the latest wave of consolidation is being driven by commercial pressures; the media intelligence sector is carrying a lot of debt and balance sheets need to be put into good order. Elsewhere many of the start-ups hatched in the first decade of the new century are coming to the end of their funding rounds and investors are looking to exit.

As well as making good on the investments of the past, the media intelligence sector is also realigning to meet the opportunities of the future. Decisions are being taken against a backdrop of technological change as artificial intelligence (AI) transforms the way media is monitored and analysed. This is not the first time the sector has been turned upside down by technology and it is unlikely to be the last.  What we are seeing with AI and machine learning is the latest iteration of a cycle of disruption and renewal that stretches back at least two decades, to the first digital revolution of 1997.

The history of media intelligence falls broadly into periods of transformation and integration.  But over the past couple of years AI has exposed a problem; monitoring technology hasn’t changed much since the heyday of the early 2000s. There have been refinements for sure, but at their heart most media monitoring solutions rely on inflexible predefined search criteria to detect mentions of their clients.

Back in the late 90s, every manual press cuttings supplier was faced with a simple choice; go digital or die. There was no legacy technology to complicate the decision. The current situation is more challenging as rules-based search methods are embedded in the big media intelligence companies. This tech is the cornerstone of their integrated service offers. Disentangling these systems in favour of AI won’t be easy. There is a lot of plumbing to replace.

As AI grows up, the technology which caused such a stir in 1997 no longer seems so relevant, and we include the UK newspaper publishers’ e-clips service in that assessment.  The market is past the point where the print edition of a newspaper offers additional value over a web page.

AI creates a point of difference for start-ups to exploit, but as the technology becomes more widespread, it might not be so easy to stay different. Over the past year many established companies have augmented their monitoring solutions with claims of ‘AI power’. Is this real change or just lipstick? It is hard to know.

AI companies could end up looking the same, at least in the eyes of their customers. Just like the Class of ‘97, the new AI vendors will fight it out to show that they are better than existing suppliers and better than each other. AI will also face the tension of maintaining a SaaS business model in the face of the PR community’s service expectations, which can be exacting.  

On the client side, the demand curve is starting to get steeper, but my sense is many buyers view AI as a drop-in replacement to meet their existing needs rather than as a means to work in a smarter way. As the experience of the social listening platforms shows, it can take time for clients to develop the maturity to exploit new tech effectively.

[Article continues below table]

What is Artificial Intelligence (AI)?

‘AI’ is an umbrella term for the ability of a computer to think and behave like a human. Among others, some of the primary goals of AI include reasoning, language processing, learning, perception and knowledge representation.

What is Machine Learning?

The terms ‘AI’ and ‘machine learning’ are often used interchangeably. But machine learning is simply one application of AI that that involves teaching a computer how to make accurate predictions when fed data. Machine learning is a subset of AI.

What are the applications of AI to media intelligence?

Some specific uses of AI within the media intelligence space include:

  • Image recognition (e.g. recognising brand logos within an Instagram image or identifying a ‘beach scene’)

  • Sentiment and emotion analysis (e.g. that news content is ‘positive’ or that a tweet expresses ‘disappointment’)

  • Content categorisation and clustering (e.g. recognising that a piece of content is about ‘Finance’, ‘Retail’ or ‘Fashion’)

  • Automated content coding (e.g. training a classifier to code similar content in the same way)

  • Disambiguation (e.g. knowing that a blog post is about ‘Apple’ the company rather than ‘apple’ the fruit)

  • Entity recognition (e.g. extracting all the people, places and companies from a news article and, in some applications, knowing they are related to each other in some way)


As the big wheels keep turning, the cycle of change is accelerating. The assets now owned by Cision and Kantar were twenty years in the making; Brandwatch has matured into a global leader in social media listening in half that time. AI news intelligence provider Signal Media is on course to cause significant disruption within its first five years.

Standing still is not an option. The next decade belongs to companies who make the most of the new monitoring technology and drive integration with the wider range of martech tools. This isn’t simply because the technology is there to use. The imperative comes from the growing expectation of customers that the benefits of innovation will be delivered.

In terms of their approach to meeting this challenge, Signal and Cision are stellar opposites. Signal is all about the monitoring tech, whereas Cision has bet the farm on  integration with its Communications Cloud. Somewhere in the middle of these extremes sits Trendkite, with a well-positioned combination of AI monitoring and martech. Kantar does not have an AI story to tell, but it does have research expertise on its side. Kantar also possesses the strongest set of tools for measuring paid media alongside the earned outputs of PR. A lot will depend on whether the new owners of Kantar push the company’s assets closer together or pull them apart.

Bending with the wind of change is something good organisations do. As well as bending, they find something to grab hold of.  The merger of Brandwatch and Crimson Hexagon is an example of two competitors realising they can make more headway together. The combination of Crimson’s machine-learning with Brandwatch’s analytics opens up a new world of opportunities to process all sorts of information beyond social media. 

This last point is, we think, the key to what the future looks like.  AI is a universal technology which will quickly exceed the boundaries of media monitoring.

There will be more processing of everything else; legal documents, public archives, medical records…anything with a font. Signal Media is already alive to such possibilities and outside of the media intelligence sector We Predict is using AI to trawl through millions of car warranty documents to help clients optimise their policy and pricing strategies.

The potential for monitoring technology to diversify beyond media is the big difference between the transformation we are seeing today and previous periods of change. For now, there will be intense competition for PR monitoring spend, but the smart players know the ultimate prize on offer is much bigger than that.


Swipe or click through the timeline below to see how the industry has evolved since 1995. We have added some observations on the last twenty years below.


To pick out some key dates from the media intelligence timeline, We’d go for 1997-2006, when the press cuttings industry experienced the Big Bang of digital transformation and SaaS first entered the media intelligence space, and the period between 2007-2014, when social media exploded and suppliers got serious about service integration.  The catalyst for each cycle was new technology; the endgame was sector restructuring and consolidation.


When transformation occurs at the initial stage of production it has an impact on all subsequent staging posts in the value chain. So it was in the late 1990s, when press cuttings agencies ditched scissors and photocopiers in favour of scanners and OCR [Optical Character Recognition] systems. Workflow was rebuilt around technology rather than large teams of human readers. Press cuttings companies could add new workloads without having to invest in more people.  In short, digital created scale.

Digital redefined the landscape as companies equipped with the new technology bought up their less advanced rivals. The model was simple; transfer the clients to digital production systems and shed the cost of the reading and cutting teams. Pass some of the cost benefit back to the client but keep most of it for further acquisitions.

The growing revenues and profitability of the sector made media monitoring agencies a credible investment target. The high-point for deals came in 2006, when industry leaders Precise and Durrants were sold to private equity houses for a combined consideration of over £100M. The entry of external investors marked a significant departure for the sector, which had previously funded its activities through a mix of earnings and overdrafts. Ownership became more short-term as investors entered and exited, plus there was greater pressure to deliver returns, which ultimately restricted the sector’s ability to focus on innovation.


Whilst the press cuttings agencies rushed to embrace digital technology a new breed of information aggregators started to enter the market, led by Moreover and Factiva. These companies sourced content straight from the newspaper publishers as digital text files, rather than scanning the printed editions.

A joint-venture between Dow Jones and Reuters, Factiva added a user interface that allowed customers to search its news feed for any name or topic. This sounds simple now, but at the time most monitoring companies worked to predetermined lists of keywords. Customers only received the content they asked a monitoring agency to look for. Factiva made it possible to search for company names or emerging issues on the fly. Factiva also offered ‘all-you-can-eat’ pricing and a much more straightforward approach to user licensing.

Armed with so many advantages, Factiva seemed the challenger most likely to disrupt the media intelligence sector but the company’s trajectory was thrown off course as it was first bought out by Dow Jones and then sold to NewsCorp. The new owners believed Factiva had a higher calling as an enterprise-level solution and switched focus away from serving the PR market. 

There are lessons the new breed of SaaS providers can learn from Factiva’s entry into the media intelligence bear-pit.  The Factiva service wasn’t especially responsive to individual customer requirements and there was no easy means of recourse if coverage was missing. To get the best from Factiva required users to invest time in understanding how the technology worked; time PR agencies did not have. Lastly, because Factiva was a global solution, users were never convinced that their home markets were being covered in sufficient depth. In the UK, Factiva’s Achilles’ heel was magazines, which were still a big part of the PR armoury in the mid-2000s.


Having spent the first half of the decade transforming and scaling their monitoring operations, from 2007 the media monitoring agencies moved their attention downstream. Durrants and Precise kick-started a new era of integration to support their clients with media contact databases, press release networks and campaign evaluation.

Precise had scaled quickly in the first half of the decade but in the years of integration that followed, Durrants probably played its hand better than anyone else; media analysis firm Metrica was acquired in 2009, followed in 2010 by the media contacts database Gorkana, from which the enlarged group took its new name.

Just as the digital transformation cycle had led to the 2006 round of acquisitions, the integration cycle climaxed in a wave of deals in 2014 led by US investment firm GTCR, which in the space of little over a year acquired both Cision and Gorkana Group as part of its plan to build a global portfolio of media intelligence assets. Kantar took Precise to add to its previous acquisitions of TNS and PressIndex.


The rise of social media introduced a new wave of social listening platforms (SLPs) to the media intelligence sector; Radian6, Brandwatch, Sysomos, Synthesio and Crimson Hexagon. Of the many, many new names which appeared from 2005 onwards, these companies deserve a shout-out because they survived the mass attrition amongst early social listening start-ups.

Backed by patient growth-investors, the social media listening platforms could do what the big media monitoring agencies could no longer do; they could invest millions of pounds in new technology. The SLPs found themselves in a playground with no bullies.

Media monitoring companies wanted in on the social media game but they knew their core technology was not equipped to deal with vast, unstructured data sets. So they agreed supply deals with the best of the SLP start-ups. It was a pragmatic move based on a belief that social media listening could run in parallel with mainstream media monitoring. The relationship worked well for a decade or so, but now the distinction between mainstream and social is disappearing fast, and the monitoring companies and social listening platforms are rapidly encroaching on each others’ patch. The SLPs are using the same technology developed to analyse social media at scale to process mainstream media content. Within the next 18 months, expect to see media monitoring companies, AI start-ups and the SLPs competing in the same space.

Thank you for visiting our blog. The team who write these posts are available to support your media insight programmes. We track and analyse media and create well-crafted, insightful reports that bring data to life. All services are available via direct or white label agreements. Please contact us at

Have your say