Walter Hale considers the speed at which artificial intelligence is developing and at the argument for PSPs to get in on the act sooner rather than later.
If Alexa, Amazon’s ‘intelligent personal assistant’ can find the right music for a family celebration, couldn’t it also tell you where your print job is?
It’s possible, but not that likely, that print service providers could station Alexas across the premises so they could keep constantly up to date. It’s more likely that, over the next decade, printers will look to automate everything they can, which will mean making judicious use of artificial intelligence.
You may already be sick to death of hearing and/or reading about artificial intelligence. That is understandable. As with all such technological revolutions, the hype often exceeds reality - especially as the technology is still developing. There was a case recently of the media hailing a new AI product that hadn’t even been beta tested as if it was going to be on sale in a few weeks. Some suppliers have over-claimed, as they are wont to do. And some experts are making completely unrealistic projections about what will be available, when and for how much. Yet we should not let fake news about ‘vapourware’ blind us to the transformative potential ahead. Nor, as we contemplate how AI could help companies reinvent themselves, should we be focusing entirely on robots and chatbots, though they will play a part in the revolution.
Robots make for great photo opportunities but the real power of AI in wide-format and digital print may lie in software’s ability to take on many chores, inform decisions and iron out a few wrinkles in the automation of the production process.
The first step on the journey to automation for a print service provider is looking at all the software you have, deciding if you are making effective use of it, discovering whether there are features you should be using but aren’t and whether it is actually improving the process as intended when you bought it. If it isn’t, it may be because the software doesn’t do what it said on the tin, or that staff - preferring to stick to the old ways - have found a convenient workaround.
Then look at every aspect of your business - from how you take orders to how you print - and determine where and why your staff are intervening. Ideally, you want as many transactions to have no touches, or very few touches, as possible because every intervention costs you money. With the advent of artificial intelligence, the prospect of eliminating touches from many print jobs – if not all of them – has gone from a pipedream to a realistic possibility in the short to medium term.
AI could be the spark that ignites the market for Web-to-print, a potentially revolutionary sales channel that has, so far, not lived up to the hype. As workflow expert Pat McGrew wrote in a blog: “Instead of sales coming in via relationships, who know your capabilities, there are jobs coming in from customers who may not have paid attention to your website and may not have followed your instructions. While most Web-to-print solutions allow for rules-based interrogation of the incoming job, in many cases, contrary to how the systems are intended to work, someone is assigned to look at the jobs and determine if they can move onto production.”
Applying AI-driven software to this problem could help Web-to-print perform as it is supposed to. It could also help companies manage variable data printing which, as McGrew notes, is often blamed for production delays, even though the root of the problem may lie in management’s failure to understand how these jobs would affect workflow.
Yet AI is not a magic wand printers can wave to make problems disappear. It is a new technology that companies have to think hard about before they proceed. As David Neely, managing director, technology and process enablement at KPMG in the US, said in the company’s recent report ‘Think Like A Start-up’: “Companies have to think strategically about AI. How do they want to leverage these technologies across the business? Where is the value? How are they going to organise to support this? What resources are required? And what are the new processes and governance you need to put in place? You need to think through these issues before you build a technology that can have a significant impact on your business.”
Neely’s point is particularly critical for the printing industry, which has too often been distracted by sexy new technology rather than making a cold, hard, business case for an investment. Yet, as he adds: “The value proposition for this technology, essentially applications that carry out a business process just the way a human does, is pretty straightforward - it offers businesses the ability to free employees from repetitive, transactional activities and let them focus on higher added-value work.”
Ordering supplies, processing invoices, diagnosing when machines need maintenance are just some of the tasks AI could take over in the medium term. With the time regained, your staff can do some of the productive things they used to do – having conversations with customers that are not about an order, getting closer to suppliers so you have a clearer idea of the innovations in the pipeline – that benefit the business even if their impact is not always easy to quantify. One significant benefit of AI is that, used wisely, it could help companies grow their capacity- and handle more transactions – without increasing their head count, a step many companies are reluctant to take in such an uncertain marketplace.
The other compelling reason for exploring, investing in or using AI is that your customers already are! In the UK retail sector, Ocado and Morrisons are already using AI and robotics to good effect. In Sweden, a rural grocery shop, which is open 24/7, has no in-store staff and is 99% managed by an app. Elsewhere the applications range from visual searches for shoppers to match their favourite item of celebrity clothing, to brewing beer (Intelligent X) and unlocking molecules in food to fight disease (a project backed by, among others, Bono.)
The next stage for many brands, manufacturers and retailers is to extend AI throughout the company, rather than use it in particular parts of the business. Many global companies are exploring whether they can use AI to manage their supply chain networks on a day-to-day basis. The return on investment here is so compelling that this could happen in the next two to three years and become commonplace by 2025. For print service providers with global corporations as clients, the message is clear: get with the programme or lose business.
Wide-format printers that position themselves as offering artisanal solutions may be sceptical about automation but even if AI was just used in admin - as opposed to throughout the company and on the print floor - it could still make a significant difference to their performance. They may also find, as industry analyst Richard Romano suggested in a white paper on automating workflow last year, “the pressures of the marketplace will force them to rethink.”
If a grocery store in Sweden could be almost completely run by a smartphone app today, why couldn’t a wide-format print firm be managed in the same way in 2020? The technology is certainly there, but the 100% automated, untouched by human hands, model of print production may not go down that well with customers. It seems more likely that a few large corporations may consider launching their own completely automated print plants.
AI is coming, even if it not’s likely to be in the shape of Amazon’s Alexa. “If you don’t have an AI strategy, you’re gong to die.” That’s what David Wenig, president and CEO of e-commerce platform eBay, told companies at a retail industry conference in March this year. He may be overstating to make his point but it would be a brave managing director who bet on Wenig being wrong.