By far the highlight of Day 2 of the 2018 Food Safety Summit (FSS) in Rosemont, IL was a presentation by Carletta Ooton, Amazon VP, Health & Safety, Sustainability, Security & Compliance – a job title she referred to as a “gigantic hat to wear” – on a topic equally as verbose: ‘Amazon’s Approach to Innovation and What It Means for Food Safety’.
Carletta, whose Twitter bio refers to her simply as a brand ambassador of Amazon, brought energy and enthusiasm to a lineup of FSS speakers who frequently lacked the ability to grab a crowd. Despite the 9:00 am presentation time she placed a can of Diet Coke on a lecturn and hit the ground running. First, a review of Amazon’s many "shopping experiences" and order-to-delivery concepts, private label products, a "treasure truck" offering mobile flash sales that she earnestly claimed was “spreading joy for everyone with a smartphone and an appetite for fun”, and the "just walk out" technology behind the Amazon Go concept store that promises ‘No lines / No checkout / No, seriously’.
This chunk of the presentation was self-serving, of course, but when representing the one company literally revolutionizing retail in the US and forcing all competitors to take heed you’ve earned the right to preen. Even as she moved on to descriptions of Amazon’s leadership principles, Carletta eschewed cockiness for a seeming appreciation of working in an organisation that follows through on beliefs it makes no secret of keeping. According to her, ‘Customer Obsession ("leaders start with the customer and work backwards") is Amazon’s most important principle, followed by ‘Invent and Simplify’ ("as we do new things, we accept that we may be misunderstood for long periods of time").
Carletta then displayed what must be a required feature of all Amazon presentations: the 'flywheel’ famously created by company founder Jeff Bezos on a napkin. Referred to by Carletta as "our virtuous cycle", it reinforces Amazon’s belief that customers always want low prices (universal) and "trust reduces friction" (not so universal).
More Bezos managerial wisdom followed: 'If you can't feed a team with two pizzas, the team's too large.' Again ... an almost painfully simple philosophy but when plugged by a global behemoth like Amazon, you scribble it down furiously, which is what everyone in the conference room (and, obviously, me) did.
Of great interest to the manufacturers and food producers in the room was Carletta's three use cases: 1) physical process control, which entailed walking through the myriad, automated steps associated with Amazon Fresh deli slicing, 2) detection and assessment, in which Carletta referred to "customer signals" as gifts and reported that Amazon receives "16 million pieces of feedback... every ... week." She then described the company's remarkable "natural language processing" and machine learning, two things of great interest to organisations seeking to use AI to gain market insights of all kinds. Finally, 3) recalls, in which Carletta explained how Amazon's early detection actually saw the company pull products from its website well before manufacturers took the same action. A fascinating glimpse of the growing power of the world's third-largest retailer.
If we’d been a bit more wary, the Sumfood team may have detected a certain lack of precision in the lengthy title of the next session – ‘How Can Big Data and Social Media Be Used In Food Safety Communications and Management?’ – that revealed itself in the session itself. A simple Google search turns up countless articles on the potential of technology to revolutionize food safety, but none of that dazzling potential was on display during the session.
First up was Patrick Quade from iwaspoisoned.com (IWP), a website that’s gained enviable press coverage in the past couple of months. Media is always seeking the 'new', and IWP is the first crowdsource entity to gain traction in the food safety arena, even if there's an element of bloodsport to its simplistic brand name and methodology. Feel sick? Think it may have been something you ate at a restaurant? Go to their site and report it. No need for a doctor’s diagnosis, no need to verify your illness was in fact caused by the food of a particular establishment … just go to their site and broadcast to their online community (which numbers 28K, according to Patrick) what you believe to be true, in as much detail as possible.
Patrick’s presentation technique did IWP no favors, as he seemed to be justifying his company’s methods even before getting roasted by questions from an audience sensitive to the damage of unverified online attacks. Hence, several forms of the same question: Is anything IWP posts on its site verifiable? Patrick sounded like a drowning man holding onto a string of words to keep him afloat: “moderators review each report”; “a combination of addresses, geo locations and contact details make gaming the system difficult”; “the way to think about our data is its signal strength”; “there’s a lot of pieces to it”; “we’re a public health organization”; “we don’t rank restaurants – we’re not concluding anything”; “we’re not judge and jury”. He offered up medleys of these phrases each time he was challenged on IWP’s methodology. Sadly for him, ‘cringeworthy’ best described his protestations.
As IWP fell under neither ‘Big Data’ nor ‘Social Media’, it was a relief that the next speaker, folksy and genial Craig Wilson of Costco, could at least boast of having big data at his fingertips – even if that ‘big data’ had nothing to do with the basic Google definition of the term: ‘extremely large data sets that may be analysed computationally to reveal patterns, trends, and associations, especially relating to human behaviour and interactions’. For Craig, ‘big data’ was Costco data: 95 million members, $140 billion company, 240K employees. Astounding numbers, but not ‘big data’. Unlike Patrick, whose already handicapped presentation was gravely wounded by his delivery, Craig is a charming speaker, and so the room listened carefully as he gave what amounted to a Costco shareholder’s speech.
It just had nothing to do with big data.