Artificial Intelligence in the DAM world

By David Howlett –

It’s fair to say there’s a difference between what most of us do every day in marketing and what the many articles about Artificial Intelligence suggest we should be doing. It’s the same as any revolutionary technological advance; the blast of hot air that you feel on the platform before the underground train arrives. It’s just taking a little longer than usual for the train to arrive in the station in the case of AI within marketing.

Two sessions covered this issue at Henry Stewart’s DAM EU Conference in London; a panel discussion followed by a case study looking into one particular Sky application during the Royal Wedding. Talking to people about AI, it’s a frustrating area. We know it’s arriving. We all see and experience its progress as customers using Intelligent SatNav apps, or Bot-powered customer service on Twitter or Facebook Messenger. But its applications within an enterprise business to business context are just another component in the ever-changing world of Martech, with companies flip-flopping on a never-ending journey from integrating best in class technology (think HiFi separates and all the cabling that involved) to favouring the one stop shop option from one of your tech vendors, with everything in the one neat box. Flick the power switch and away you go. It’s that easy.

Back to DAM. We are aware of progress in three distinct areas where AI has been considered.

Autotagging of imagery and assets

This uses image detection software, and is available from many vendors (e.g. Google, Microsoft) and increasingly via DAM software vendors directly, many of whom have integrated the ‘engine’ as a module within their own systems.

In relative terms it’s a very well developed technology, but it’s put to the test pretty severely in a DAM context.  See our whitepaper for more detail.  The bottom line is that it’s useful during a large migration activity after a certain amount of training on your styles of imagery, and / or when you are considering a long term solution for continuing high levels of asset upload.

User support, powered by ChatBots

This uses well proven technology commonly used in customer service solutions that can be re-purposed to provide an easy to use 24/7 interface to match user queries to a library of answers within a closed network of employees and agency partners. Technology itself is the easy part.

The challenge is (1) understanding the range of questions you are going to be asked, and (2) having a definitive knowledge base of responses matched to those questions. This is a challenging undertaking for many organisations, but once you have this, then ChatBot deployment makes sense, and payback will be assured over time in larger enterprises. Your DAM people (whether internal or external) are freed up to spend more time on change management, communications activity and liaison with regional and brand teams to drive improved DAM usage.

Our whitepaper explains how this might best be prepared for within the DAM and wider Martech ecosystem.

And finally, there are a raft of specific instances where AI techniques can help clients overcome specific challenges by using ‘Big Data’ combined with machine learning, to guide clients to the best solution. One such example, quoted by Thomas Eusterholz on the Panel was the integration of PIM data (Product Information System) by flagging potential errors. We also note reports that the use of copywriting Bots is now mainstream on e-commerce platforms, with their ability to propose product ‘description and benefits’ text during the sales process using auto-generated content.

Our Panel hopefully demonstrated the very diverse points of view to AI by the various stakeholders involved in DAM.

• DAM budget holders and business owners see the potential for cost savings and technology use, but will not want to risk choosing an under-developed tech solution.
• IT stakeholders will be pleased to be involved, but will be keen to understand the viability and cost-effectiveness of every solution before implementation.
• The current DAM team will often be wary of an under-developed tech solution and how that might adversely affect DAM usage, as well as being personally nervous about the changing shape of their own responsibilities.
• The DAM user base will be driven purely by the improvement over the past system. How quickly can the issues be resolved? If it’s not broken, why fix it?
  Take these into account when introducing projects into your organisation, and always set up multi-functional projects so that the issues can be discussed early on and resolved before a launch phase. AI is only going to grow alongside its older cousin, Automation, so it’s best to be prepared and seek out the right time to start advocating its use according to your own, specific requirements.

David Howlett
Marketing Director

Victor Lebon (ICP CEO EMEA & APAC) moderated the Panel “Hype or Hope – Can Artificial Intelligence Help Power DAM User Adoption and Engagement?” at DAMEU 2018. The Panel comprised:

• Maria Efstathiou, WW Business DAM Product Owner, Ubisoft Entertainment
• Thomas Eusterholz, Co-founder, AI Integrations
• Kristina Herz, Head of Digital Asset Management, Royal Horticultural Society
• Romney Whitehead, Digital Content Specialist and DAM Consultant