Fixing supermarket self checkout

Buying your groceries at a supermarket is pretty simple – wait in a queue, hand your items over, pay for them, pick them up and go.

However the checkout counters are expensive, take up valuable space and only work when there are staff present. That makes it difficult to ramp up or down for demand, and, queues are generally required to make the costs economic.

Enter the supermarket self-checkout system. These take up a lot less space, but allow the shopper to avoid the queues and check out at a speed that they can control.

However it’s not so simple. The systems bought into lace, at least in NZ, are appallingly complex. From my experience the process is something like this:

The clearest demonstration of failure is at beginning, with the queue. An efficient checkout system would be able to take up all of the demand and remove the need for queues, as the Air New Zealand check-in kiosks do at local airports.

The experience of self-check out is vastly inferior to the traditional approach, especially given that almost all checkout operators I encounter are interesting, efficient and genuine. I generally prefer to avoid human interactions when making purchases, but the current state of play is just woeful.

The answer lies in trust. The supermarkets clearly don’t trust their customers, and the asinine scanning and weighing procedures slow the entire process to a crawl. The answer is to turn the tables, and place the onus of getting he checkout right on the customer, rather than trying to solve a potential issue with technology. Almost everyone is honest, especially in New Zealand, and a few tricks can increase compliance to essentially 100%.

Firstly, install cameras and discrete signs on the checkouts noting their use, and noting that failing to check something out is deemed as theft. Make the design of the kiosks open enough so that the customers can be observed from several directions by other shoppers and staff.

Secondly bring in a system of random audits, picking one in every 100, say, and doing so using the checkout itself rather than human selection. The audit will involve a team of two nearby who will check he contents of the bag against the receipt after checkout, and then give a coupon, discount or other small gift to everyone for their compliance. Treat non compliance as a mistake, unless obvious for high value items, but flag their credit card for future increased audits or potential banning from self-checkout.

Finally remove cash as an option for self checkout, which will remove the complex and breakable cash-acceptance equipment, and encourage the continuous transition to debit and credit cards. Accepting only plastic will also give the ability to retain more control, as well as offering the ability to link credit cards to loyalty cards and provide benefits.

Other options

Why not for now allow us to scan items as we shop, using an iPhone application or a portable scanner attached to the trolley. That we we can arrive at the checkout with the items already in a list, and perhaps we can even pay for them with our phone or with a credit card acceptor attached to the trolley.

Ultimately we should be able to lose the checkout process entirely, as each product can have an RFID tag, and your phone or wallet can contain a remote payment mechanism which is activated as you leave.

About Lance Wiggs

@lancewiggs
This entry was posted in NZ Business and tagged . Bookmark the permalink.

10 Responses to Fixing supermarket self checkout

  1. Ben says:

    The system you propose actually exists, but is pitched at a full trolley shoppers rather than the ’12 items or less’ crowd. Pak N Save has Shop N Go (http://paknsavekilbirnie.co.nz/40/shop_n_go). You take a portable scanner around, scan your own items, then pay a person at dedicated checkouts. They have random checks to reduce theft and even treat minor non-compliance as a mistake. It would probably be a small leap to extend the system to top-up shoppers. Maybe their wireless scanner system is so crazy expensive they aren’t willing to scale up for the large number of extra users that would entail?

    • Lance Wiggs says:

      That sounds great. I guess it shows that I don’t shop at Pak N Save. We generally walk, cycle or at as pinch motorcycle to the supermarket, so buying in bulk creates a real transport issue. It says something that Pak N Save, with larger carts, very low margins and a lower socio-economic target market, has done this well while the more up market stores have not. Pak N Save has always been about self-service, but the others seem to be missing the point that self-service, done well, is better than full service.

    • Su Yin says:

      Great initiative but you have to apply for a scanner (‘approved applications take between 14-21 days’) and undergo training before they will let you have one.

      Y U NO TRUST ME?

  2. Peter Crow says:

    Hi Lance
    Our experience is quite different from yours.

    We use self-service checkouts by preference when buying a few items. I’m surprised you make the self-check process out to be so complex because that is not our experience (at New World and at Pak-n-Save). To me, the process seems to be exactly the same, except that I am my own checkout operator (and I’m no a patch on the professional ones). The professionals scan, weigh (if needed), pack, wait for approval for wine is required, etc etc.

    Regarding queuing. Pak-n-Save Porirua has approximately 9 self-checkouts and NW Porirua has about 5 five memory. One queue feeding in and you go to the first available unit. Queueing for more than 30secs is unusual in my experience.

    Given all this, I wonder whether your self-check argument is a red herring, and that the main event is more a call for a radically different approach to buying things at supermarkets? And as Ben states, the Shop-n-Go option seems to work well for large shopping expeditions.

    Cheers -prc.

  3. David (@HolmesPD) says:

    Approximately 2001 in Sydney, we used the services of an internet supermarket for several years. Eventually we were part of a trial, where we had a basic portable scanner and scanned whatever items we wanted to purchase (from our pantry or as we ran out) and then downloaded the items to our computer, and added additional items over their website. Delivery was $6 or $8, and we could select the delivery time in 3 hour time slots up to 11pm at night. My point, lets think laterally, do we need supermarkets, can we have engineer a different system. Smaller supermarkets + internet ordering. The supermarket operators can use a central warehouse, with cheaper fit outs but a distribution system is needed. possible……. Probably!

    PS: Their tag line……… “Get your bananas in pajamas”

  4. My main issue with the self service checkouts is that you need ‘assistance’ with Amex and booze. I would rather they tie my Airpoints card with the my CC card and a once off age verification so I don’t have to worry about signature checking and age verification. Once that is in place the number of steps from getting to the counter to leaving is massively reduced.

  5. Nick Bowmast says:

    You’re so very right about the trust issue, which exists in every self-service scenario – customers trusting the technology being the initial leap.
    I ran some design research for a major UK supermarket.

    http://www.bowmast.com/case-studies/#6

    Supermarkets are geared towards promotion and impulse buys. These self-service technologies come into their own when they help shoppers avoid these, and stick to their budget.
    Compared to in-store, online services offer the customer greater control over their spend as the total is permanently visible, and rising with each item added. For the same reason, the handheld scanners would likely result in a lower total spend.
    The question here is… Does a customer become loyal to the supermarket because it helps them to spend less? .. or does the transparency of the total spend take away the guilty pleasure of adding those impulse buys to the basket?

  6. bwooce says:

    I think you’re overlooking that this is FMCG. High volume, but low profit per item. A 2% fraud rate on this would kill them and there are many types of subtle fraud that can occur (non-scanning, substitution, etc)

    Additionally, being pulled out of line sounds much like the TSA experience at a USA airport. I don’t think I’d go back after that happened, irrespective of how nice they were. What if I was in a hurry? That’s why I do the work myself in the first place…

    See a recent public example of how there is a klepto hidden in many of us: http://www.dailymail.co.uk/news/article-2084067/Antony-Worrall-Thompson-sorry-shoplifting-cheese-wine-Tesco.html

    I suspect they actually have hidden cameras over the terminals already here in Australia. I base this on the somewhat random apperance and clearance of “please wait for a person” messages when I take bags off the scale (cos they’re full) and admit to this “crime”. No discernable pattern exists, although perhaps it’s just random on purpose :-)

  7. Bart says:

    This looks like a workable alternative:

    http://www.wimp.com/marketingcampaign/

  8. Microsoft kinect got also inserted into a shopping cart as well, scanning item automatically as you place them in the cart, and simply paying by credit card automatically when you go out through the doors. It also recognize when you remove item from the cart – see: http://www.popsci.com/technology/article/2012-03/video-smart-shopping-cart-future-follows-you-through-store

Comments are closed.