Amazon sneezed. Are you scared?

So Amazon casually announces, what is tantamount to "associate-less" or "fewer associates" based store and the rest of the retail industry goes crazy! :) Amazon has to be complimented in how it continues to think outside the box, and pushes the envelope, forcing retailers & technology companies to scramble, even if some of it is marketing hype initially.

Most of us can agree, that the underpinning technologies have existed for a while, even if we peddle them today with newer fancier monikers, from IoT, Image/Video recognition, to Machine Learning,

IoT is all about doing something useful with things that are connected, but its roots can be traced back to the old programming logic controller (PLC) days when the Allen-Bradleys of the world ruled industrial and logistics automation. Even today, the industrial protocols like Zigbee & Thread are still very common and prevalent and frankly far more superior in their mesh networking than the ubiquitous 802.11x found in our Wifi.

Image/Video recognition, over the past 10+ years, has been piloted and applied selectively to combat concepts like "sweethearting"; where a cheaper product is scanned in lieu of what's actually taken or "scan action" is faked.

Artificial Intelligence surged back in the 80's with Prolog, LISP and then got relegated to computer scientists in academia since the tools, programming sophistication and computing power was just not there yet.

Amazon has cleverly combined multiple nascent and established technologies to create a new shopping experience, but is this broadly viable in the retail industry?

Well, don't hold your breath with the hope that your supermarket or department store will sport this anytime soon! My initial reaction is that whilst certain types of retail formats can certainly benefit; think smaller-formats, identifiable-customers, limited set of actively shopping customers at any given time, regulated environments such as airports etc... such a solution in the near term; has got to be financially unjustifiable and operationally untenable.

Of course, Amazon may simply be using this to test out the viability of the technology, hone its algorithms and turn around and sell it as a turnkey solution to retailers, all with the streaming data going right into its cloud for analytics. In which case, good on them, and bring it on! Lack of competition breeds mediocrity, and that applies to all of us, retailers and technology firms :-)

A more practical approach for most retailers would be to explore how IoT, and instrumenting their stores and combining that with mobile & image recognition can help them squeeze more productivity out of their associates, reducing power consumption, adjusting shifts to meet demand, reducing inventory carrying costs, stock-outs etc.

My company, IBM for e.g. has a very exciting program around our IoT platform where we "show, not talk", by taking over a facility, store etc.. and demonstrate how getting things connected, applying machine learning on top of their generated telemetry can be used to deduce actional insights.


Technology aside for a moment, a larger socio-economic issue is taking hold, with machine learning, automation, robotics etc. In a retail store, store-associate payroll is typically the largest operational expense, followed by telecom. Such a solution essentially strives to commoditize the function of a store associate and transfer some of these responsibilities to machines.

Of course, you'd still need to stock the store but we'll have robots for that, after they've unloaded the merchandise from the self-driving trucks or drones :-)

Any and all occupation, it seems, can be taken over, interpreting x-rays, diagnosing diseases, driving our cars. Heck maybe I'll send my robot to the store to chat up with Amazon's robots, and buy stuff for me... but the socio-economic question is... would I even have a job/income to pay for it! :) 

As a technologist, and a consumer, I'm of course super excited about all of this; but as a social human being... not very!

Bob Korzeniowski

Wild Card - draw me for a winning hand | Creative Problem Solver in Many Roles | Manual Software QA | Project Management | Business Analysis | Auditing | Accounting |

7y

"Amazon has to be complimented in how it continues to think outside the box" No. Amazon does not think outside of the box. It believes that employees are nothing but cost centers. That is deep inside the box thinking there. Amazon also works their people to death, and that thinking is deep inside the box. Amazon believes that employees have no value and thus must be eliminated - again deep inside the box thinking. If they actually thought outside of the box, they would believe their employees are assets and not just cost centers; enact and enforce a no layoffs policy; and actually value their employees as the reason they're getting richer. Now THAT is outside the box thinking!

Danny, Thanks for taking us through the journey of a customer at a retail store over the last 20 years I would say. I completely agree with you that perhaps transaction cost per employee has come down a lot and going through this approach will probably take out some more cost, however the question that we need to ask is how many customers today want them to be tracked throughout the store and get in to their privacy. Also the personal touch will be completely missing, why will someone then go to a store and not get something from the web. I have my serious doubts over how it will address the security concerns, to me unless the customer moves completely out of the store the billing should not be done, however if we wait for the customer to move out of the store without a check there are chances that it will pose security threats. I am curious to see how they can by just swiping the phone track the customer through all the purchase, assuming that the customer switches off the phone or comes out of the internet I do not see how tracking of customer can be done without visual data, i.e camera only focusing on someone's face all the time ? At the same time I think Amazon has generated curiosity and want to be seen as market leader..

A typically thoughtful and contrarian view Vish. You make some interesting arguments on the social impacts of this technology. To offer my own contrarian view, I'd say that the trends are clearly pointing towards the check-outless store and it's already been here for some of us for the last few years. If you rewind in the grocery industry only 30 years, you would have found 20 -25 lanes in a supermarket fully manned with associates on a Saturday afternoon, keying prices and taking cash. A typical checkout transaction took 5+ mins plus per customer and each till was 3 - 5 customers deep. 10 years later and due to ubiquitous use bar codes and reduction in cash only 1/3 of those lanes were manned even at peak, the checkout time had dropped to 2-3 minute. 5 years after that - self checkout was introduced. Now in a typical super store only 10 - 20% of lanes are even manned and one employee watches over a cluster of 4-6 checkouts. Transaction times haven't fallen, but the transactions per employee (and therefore retail labor cost) has fallen dramatically. The convenience sector in the UK has adopted this model to the extreme with ONLY self checkout available in some stores. A brief dalliance with rfid in the early 'noughties' proved the need and the benefits, but the technology always represented an additive cost per item, so was never adopted. Fast forward a couple more years and we start to see a few companies adopt customer self scan - either with their own device or a retailer provided handheld. The transaction time is now close to zero and involves Zero employees. (On a personal note I use Sam's club's 'Scan and Go' app on my iPhone every week and haven't required the assistance of an associate for over a year., other than a 15 second interaction with the exit checker. The Amazon Go trials are simply an extension of this trend. Assuming they can make it work practically (I have few doubts), the commercial pressures will no doubt prevail and this will eventually be widely adopted (in my humble opinion). There is very very little value add in having a human being transferring data to a machine in the checkout process and so it will eventually disappear. Amazon's efforts are no doubt highly proprietary and focussed on allowing them to continue their battle for volume growth in the bricks and mortar space. They need to keep the juggernaut moving in the battle against Walmart et al. It will be interesting to see if they chose to commercialize the offering in a bid to affiliate the small stores / independents and mirror their 'marketplace' approach. Personally I'd say they will keep this as their 'secret sauce' for a while yet. Anyone selling POS systems better look out though.......

To view or add a comment, sign in

More articles by Vish Ganapathy

  • The starting problem in retail

    The starting problem in retail

    Retail has a starting problem. The more I talk to clients, the more I am convinced that we have a starting problem in…

    3 Comments
  • The people & process problem with data science in retail

    The people & process problem with data science in retail

    Given that existential urgency facing retail, This mismatch between the general nods of approval of the power of data…

    5 Comments
  • The retail data problem

    The retail data problem

    At a recent innovation forum, I was seated next to a senior exec of a large, reputed and successful retailer. He…

    1 Comment
  • The elusiveness of fundamentals in retailing

    The elusiveness of fundamentals in retailing

    Every year, I try to go back in time; 2-5 years to recollect what we've been saying. It's the same thing it turns out.

    2 Comments
  • Artificial Intelligence slowly learns retail

    Artificial Intelligence slowly learns retail

    Just like that. Artificial Intelligence and Machine Learning systems are everywhere! Everyone has a framework, a…

    3 Comments
  • Sharing economy and the magic of ecosystems

    Sharing economy and the magic of ecosystems

    A few times every year, my neighborhood lays out stuff, on our driveways, so that neighbors can pay neighbors, some…

    1 Comment
  • Are retail stores like COBOL?

    Are retail stores like COBOL?

    Seriously. Are retail stores really like COBOL? Hanging around, still the heartbeat of many businesses, never really…

    9 Comments
  • Embracing (a bit of) mediocrity for success

    Embracing (a bit of) mediocrity for success

    Imagine yourself as a caribou, munching away with family & friends on the savanna grasslands. Enter a lioness who'd…

    1 Comment
  • Sprinkling some grey cells

    Sprinkling some grey cells

    A few times every year, my neighborhood lays out their stuff, so that neighbors can pay neighbors, to buy the stuff…

  • Will your ignored data bite you back?

    Will your ignored data bite you back?

    "Are your kids health, their breathing issues not worth seven thousand dollars?". The HVAC salesperson asked.

    1 Comment

Insights from the community

Others also viewed

Explore topics