A decade ago “price optimisation” was expected to transform retail operations and improve proﬁtability. Despite the hype few UK pilots progressed to roll-out, but elsewhere the technology has become mainstream. Penelope Ody asks whether it’s time for multichannel retailers to catch-up.
Back in the early 1990s a spinoff from Manchester University’s computer sciences department came up with a set of clever algorithms for comparing and adjusting petrol prices to reﬂect local competition. The major oil companies were delighted and the academics happily spun off Knowledge Support Systems as an independent and leading edge IT company. The science of “price optimisation” was born.
Fast forward to the 2000s and bofﬁns in Boston and Silicon Valley launched companies such as DemandTec, Proﬁtlogic and Khimetrics all offering very clever “black box” solutions that promised to turn retail pricing decisions from an art into a precise science, so helping to increase margins, improve markdown management and make promotions pay.
Systems basically used price elasticity –how much shoppers will pay before switching to a cheaper (or more expensive) brand – to then juggle the mix, pushing some prices up and others down to give the best margin performance. The same systems could calculate where promotional activity would cannibalise higher margin lines, with a markdown component calculating which items needed to be discounted, when and by how much, to minimise margin erosion.
With price optimisation hailed as the “next big thing”, set to transform retail operations, Oracle snafﬂed up Proﬁtlogic, SAP bought Khimetrics, IBM [IRDX VIBM}], “and we have some ongoing conversations with clients, but it is a rather more live issue in other territories than the UK.”
Itim, which acquired optimisation specialist Proﬁmetrics in 2009, has a strong customer base in South America: “Differential pricing is commonplace over there with prices optimised for different regions,” says Operations Director Martin Schoﬁeld. “Retailers need to sell as much as possible at full price early in the season and we’re seeing more interest here in how price can affect sellthrough – although differential pricing is not on many agendas. Some UK retailers seem to struggle to maintain one price ﬁle, let alone 10.”
North American retailers and many in mainland Europe, notably the French, have also opted for price optimisation while, as Kathy Beck says: “UK retailers are so far ahead of the curve in many other aspects, but not when it comes to the science of optimisation”. Performance improvements can be impressive. Revionics user, Family Dollar – a value grocery chain with average basket size of just $10 – reported gross margin improvements in optimised product categories of 4% in a presentation at NRF in January, with increases around 5% in gross margin and 10% in sell-through for its markdown optimisation pilot.
Howard Langer was an early enthusiast implementing DemandTec during his years at B&Q [IRDX RDIY]. Now Global Capability MD for pricing and promotions at Dunnhumby he also runs a retail forum under the auspices of Revenue Management and Pricing International.
“Pricing is the biggest lever any retailer has and it is at the heart of what they do; it is important to understand pricing as a management discipline – much as supply chain is now recognised,” he says. “There are plenty of tools available to help but retailers seem reluctant to use them.”
FOLLLOWING THE COMPETITION
This reluctance seems to hinge on current preoccupations with promotions and competitive pricing. Rather than establish their own pricing strategy based on target market, merchandise mix and elasticities and then holding their nerve, many retailers – both on- and ofﬂine – seem to be more concerned with pack behaviour and following the competition. Monitoring that competition has never been easier with companies like Proﬁtero delivering daily reports on web prices to dozens of retailers. “All UK supermarkets use online pricing data,” says CEO Volodymyr Pigrukh “and the US is catching up, although many there still just do competitive in-store price checks.” A single daily web scrape of competitive prices is also just the start: in the run-up to Christmas some retailers were demanding four daily updates while one of Proﬁtero’s French customers – where local pricing is the norm – monitors nine retail chains across 600 locations: some 12 million price points daily.
In North America you’d be a very unusual retailer if you didn’t have a price optimisation solution,” says Mike Taylor, DemandTec product strategy lead at IBM (UK), “whereas in the UK pricing is set by competitiveness. Thanks to web scrapes, everyone now knows everyone else’s prices and can understand their stock position. Price optimisation is no longer just a point solution. Since IBM acquired DemandTec, we’ve integrated it with range and promotional activities as part of a total merchandise offer. As the online world has taken off we’re all understanding more about customer preferences and price is part of that. We’re already seeing personalised promotions with personalised prices – not yet to a segment of one, but the segments are getting smaller.”IBM is also looking at how price changes affect channel dynamics. “Customers behave in different ways and channel elasticity varies for different products,” adds Taylor.
“It means that optimisation science is now going cross-channel and retailers can tweak prices in different channels to inﬂuence where customers shop.” SAP, too, has moved Khimetrics on. It is now the basis of the company’s performance, insight and optimisation division working with its big data analytical engine, Hana, to optimise both prices and promotions. “We’ve been looking at afﬁnity insights,” says retail industry principal, Alan Taylor, “as the technology allows you to analys billions of transactions in a few minutes and improve the promotional mix. For example, 2 litre bottles of Pepsi are almost always bought with crisps so if Pepsi is on promotion don’t discount the crisps as well as they’d be bought anyway.” In a proof of concept implementation in the UK, Taylor reports an 8% increase in proﬁ ts: “You can optimise prices of related items and linked products and there is interest in the technology from hardlines such as DIY as well as grocery,” he says. “So much gross margin is given away in promotional activity that retailers need to do something.”
JDA’s system combines strategic pricing, promotional optimisation and markdown optimisation as “price lifecycle management”. “Pricing is really part of the planning process,”says Lee Gill, “but in many retail businesses it is marketing that makes the pricing decisions not the buying department. The buyers must buy to a price and then store operations deal with the execution so there are silos and no single custodian of pricing strategy.”
Howard Langer agrees: “Where retailers have pricing teams they tend to be focused on competitive pricing, but pricing should really be a strategic issue related to customer targeting. Understanding customers is what Dunnhumby has been doing for 25 years so combining KSS pricing technology with Dunnhumby’s customer insights means that you can combine customer engagement and loyalty with price sensitivities. There is no point in promoting products to non-proﬁcustomers or optimising prices to push volume to the cherry pickers.”
Price optimisation is generally regarded as being most effective for fast-moving, replenishable merchandise– typically groceries, DIY, sports goods, consumer electronics and so on. Fashion tends to be short-lived and often own-label, making the markdown component of the pricing lifecycle more relevant. There are also cross-channel implications that can affect whether online and in-store prices should be the same or whether local pricing viable.
“Operations are more expensive in-store so you’d expect higher prices,” says Mike O’Hanlon. “Some retailers make KVIs (known value items) the same across all channels but take a more ﬂ exible approach to other lines.” Matching price to location is just as easy: “The prices shown online can be varied by zipcode to match that in local stores,” says Kathy Beck, “and you can always empower store associates to match online prices if they’re lower – although it seems that despite the hype very few store shoppers ask for price matching.”
For many retailers blanket markdowns and relentless competitive price-matching is already playing havoc with margins – and we all know what happens to businesses that don’t make a proﬁt.