The Drum Awards for Marketing - Extended Deadline

-d -h -min -sec

Artificial Intelligence Machine Learning Technology

Before starting an AI project, brands need a 'shiny object syndrome' diagnosis

By Daniel Henriksen, Technology and Operations Director, Programmatic

August 1, 2017 | 8 min read

All this week Daniel Henriksen, technology and operations director, programmatic at OmnicomMediaGroup in Asia Pacific, is helping marketers diagnose whether they have shiny object syndrome, when it comes to data and analytics. The four aspects brands need to consider are reality, organization, people and process.

diagnosis

Artificial intelligence and machine learning is the promise of using data and technology to build intelligent agents.

This, potentially, can transform companies and organizations into to high-performing ecosystems through logical autonomous decision-making, maximizing the chance of success towards a defined goal or objective. Theoretically, this could all be done by humans as well, but obviously not with the same speed, precision, and processing power that machine computation possesses.

So, the answer seems simple: Start AI projects ASAP!

But, is it that simple? If we had to look at ourselves as individuals and consumers when it comes to making decisions about technology, it's undeniable that we suffer from "Shiny Object Syndrome (SOS)".

"It’s called shiny object syndrome because it’s the equivalent of a small child chasing after shiny objects. Once they get there and see what the object is, they immediately lose interest and start chasing the next thing. For companies, rather than literal shiny objects, it may be business objectives, marketing strategies, clients or even other business ventures"

Like individuals and consumers, companies and businesses can also suffer from "SOS", where they tend to always be chasing the next thing, instead of stopping and assessing their organizational readiness to transform and integrate modern technologies successfully.

To some extent, this seems to be the case with artificial intelligence, where everyone wants to integrate and use machine learning instantaneously within all business processes to gain immediate efficiencies and increased profitability. Other examples of "SOS" from the perspective of a media agency, are clients who wants to jump straight from ad network buys to complex automated media buys, or clients who jump straight from static jpeg banners to data-led dynamic creative with multivariate testing.

It can be a dangerous endeavor to embark on such complex journeys without ensuring that the organization is ready, and ensuring that the right people and processes are in place before doing so. Therefore, it's important to assess how data and analytics currently is being integrated into the organization, developing a strategy of how to become a data-driven company - utilizing computed data-driven decisions to improve all aspects of the business.

Where to begin?

It's always easier said than done, and there are many components impacting the development and activation of a strategy to better integrate data and analytics across all parts of an organization. According to IDC, worldwide revenues for big data and business analytics will grow to more than $203 billion in 2020. It's apparent that most organizations or companies to some degree already have planned to start this journey, and many resources are being allocated by CEOs, CIOs and CMOs to fund investments in technology across all parts of their business, ranging from intelligent supply chain to marketing automation software.

Simply acquiring the latest technology and software is great but, as described in this article published by Harvard Business Review, the significance of the people and organizational components required to integrate data and analytics is often underestimated. To improve readiness and become more equipped as an organization to transform and adapt to a world of data and technology abundance needed to run a profitable business, there are some areas which are important to assess when developing a strategy; the reality, the organization, the people and the process.

Reality

Like the science-fiction film ‘Inception’, knowing your own reality can sometimes be quite difficult, and hard to accept if one doesn't necessarily agree with that reality. The same goes for organizations when it comes to change management processes and transformation, as it is important to be realistic and very self-aware before implementing change. Being overly optimistic will only backfire, and is something that business owners and entrepreneurs sometimes suffer from - as the company is an extension of themselves. When it comes to integrating data and analytics into the core of all business processes, there are a few things worth to assess regarding overall organizational readiness:

Resources

Having infinite resources is not common for most organizations, and most organizations must make very calculated and smart decisions in terms of how to allocate the certain limited amount of resources (time, money, know-how, raw material) to their disposal every single day. One might look at Google, Facebook, Amazon and Alibaba thinking there isn’t an end to their endless deep pockets, but even they have to make decisions about resource allocation, especially if they want to keep their dominant status in their respective fields.

The same goes for resources when it comes to integrating data and analytics into each core business process, and the resources spent on these types of projects could have been allocated elsewhere. Therefore, it’s important for every organization to evaluate overall of how to allocate based on the current- and desired future state of the business. According to Gartner’s CMO Spend Survey 2016-2017, survey data suggests that marketers are expected to allocate 27% of their expense budget to technology, meaning that it seems to be an area of increased focus for most companies to invest in.

Priorities

A well-known fallacy is that we as humans or organizations think everything can be done at once, and we tend to be overoptimistic of our own capability to juggle many projects simultaneously without negatively impacting the output quality of each individual project. Organizations need to be strategic with their priorities, and decide which projects/tasks to focus on in terms of what will bring the business forward the fastest towards established short- and long-term goals.

The amount of business processes is vast, and whether we are talking procurement, production, finance or human resources, data and analytics can improve each one of these processes. However, depending on the challenges and nature of the organization (government, commerce, service), there will be a need to prioritize certain processes within the supply chain in order to allocate the right amount of resources when it comes to time, money and knowledge.

Barriers

Barriers in organizations are almost always encountered in any type of change management process across teams, workflows of operational processes. So before attempting to implement change and execute a strategy around utilizing and integrating data and analytics into core business processes, it’s essential that barriers are identified and acknowledged prior to initiating. With artificial intelligence, barriers to successfully integrating intelligent agents and machine learning might be; Lack of technology, Lack of talent/know-how, Limited organizational understanding, Data silos or simply fear of the unknown. They can all be overcome, but understanding the barriers first is key.

Goals

It goes without saying that a goal or objective needs to be associated with any type of project, task or change management process. The most dangerous part of “SOS” is that humans and organizations simply chases something because it’s “shiny”, and not because they’ve identified what value it will bring, or what problem it will solve. Getting the newest iPhone is great, but if all you use it for is texting, what does it matter that it can do 4K video recording at 30 fps? The same goes for wanting to use and integrate artificial intelligence into business processes, if all you need is simple automation to structure data to make more informed decisions for the organization, this might be achievable with more simple solutions such as Microsoft SQL or Microsoft Access. Identify the goal, and start from the current state of the organization.

The next stages; organization, people and process, will be published each day this week as part of a series on The Drum.

Daniel Henriksen is technology and operations director, programmatic at OmnicomMediaGroup

Artificial Intelligence Machine Learning Technology

More from Artificial Intelligence

View all

Trending

Industry insights

View all
Add your own content +