Lessons from a decade in the first and last mile

This blog is adapted from the keynote presentation at Leaders in Logistics Summit 2024.

Having spent a decade (which I recognize is not a massively long time in this industry!) working at the coal face of the first and last mile, I wanted to take my slot today to discuss some of the key changes we’ve seen from our unique perspective as Doddle – and to look out at the next decade through the lens of Blue Yonder, the business I now work for.

And when I think about the things we’ve really earned the right to talk about, out-of-home delivery or PUDO is definitely one that we’ve developed a real focus on and a library of content about. It’s been a story of huge growth in the last decade in this space.

In the UK, major carriers have doubled their networks, other providers have exploded into the space powered by lockers, PUDOs or increasingly a mixture of both.

In Europe, Geopost went from circa 20,000 OOH locations in 2015, to 58,000 in 2020, and then reached 100,000 locations in 2023.

But it’s not just out-of-home that has rapidly peaked in this time. Delivery speeds have also ramped up. It’s hard to imagine today, but Amazon’s initial ‘free delivery’ proposition started at 9 days. Today it’s down to 2 days and in some cases quicker than that. The average number of days for a US ecommerce delivery to arrive has dropped significantly even since 2020, from 6 days to below 4.

That speed isn’t just driven by transportation time though – it’s also about much faster processes further up the chain. Accenture reporting says that Amazon used to take 18 hours to move an item through a fulfilment center into the right truck for shipment, back in the early 2000s. Today that time is around 2 hours, which illustrates the trend for faster and faster processes.

E-commerce is coming off the back of pretty much uninterrupted growth at ~10% every year (in the UK), and after the spike and fallback during and after the pandemic, we’re now in a position where the forecasts call for growth at <5% for a few years, before falling back to relative stability, in line with overall retail.

That suggests that post-pandemic, we’ve reached a state where the market is more mature and stable, but the years of almost automatic growth are probably largely gone. The question is what does that mean for the industry?

It has massive implications, because that growth was funding growth in other areas, areas which naturally lag behind ecommerce growth but are intrinsically linked to it.

So what happens to OOH, and what happens to the desperate drive for speed?

Well for out-of-home, looking at an adoption S-curve, you would expect that we will reach the laggards stage, and OOH location numbers will stabilize and that growth will slow down.

The point is that the gains become more and more marginal. InPost set the goal of 10,000 lockers in the UK, a headline number that it believed was what was needed to achieve success in this market. But that was due in 2023, and today they have around 6,000 lockers – still an enormous achievement and driving change in the market, but the marginal improvement between 6,000 and 10,000 is a lot lower than between 2,000 and 6,000.

The same applies to speed. Once you’re Amazon and processing items in under 2 hours, there’s less time to gain, and the cost of doing so increases.

The past decade has been about building, developing infrastructure, technology, capabilities – whether that’s 100,000 out-of-home locations, 10,000 lockers, the ability to move items through fulfilment inside 2 hours, to deliver them in 2 days, or 1 day.

The next decade is about optimizing that infrastructure, technology, and capacity. It will be a transition from form and function (building stuff) to efficiency and utilization (making the most of it).

This is where our thinking as Doddle has changed since our acquisition last year. Now we have the perspective of Blue Yonder – a business 40 years in the making, not 10, and with 6,000+ people, not 90. The horizon now goes out much further, and where we used to think about planning in terms of 12 weeks, maybe 12 months, we now have the capacity to look much further ahead, and much wider. Not just next year but the next decade. Not just the last mile or out-of-home, but across a supply chain. Not just optimizing single functions, but orchestrating across functions to make a much bigger difference.

As Blue Yonder, we think about three core trends.

Best-in-class technology is fairly self-explanatory. The tools you need to do the job, in a complex and changing landscape. They have to be proven, they have to be effective, and drive return on your investment many times over.

Interoperability is more interesting. The basic premise is that interoperability is about synchronizing data, workflows and systems to improve decision-making and create transparency. It helps businesses move away from complex, manual processes, and from silos that restrict what’s possible to achieve.

To think again about the start of the decade of Doddle, we had major frustrations because while we had amazing stores with a great customer experience, +80 NPS, great staff workflows and the potential to transact a lot of volume. But we couldn’t get carriers to price in the consolidation benefits of out-of-home delivery, and so we couldn’t get priority or promotion from retailers to drive volume.

Now if the carrier pricing system was able to talk to the retail order management and carrier management systems, and those systems could connect to our stores and capacity, it would have been a no-brainer to utilize that opportunity to consolidate – the systems would have made it clear and obvious when there was capacity, how valuable it would be, and what the right price to offer would be.

There’s a similar problem with returns, which has been an issue since the start of e-commerce. It’s a complex problem of customer data, order data, warehouse data and transport data. But because big businesses haven’t been able to connect the dots on these parts of their supply chain, they’re restricted to quite blunt tools – like swinging from free returns for all customers to charging all customers for all returns. It reflects systems which aren’t interoperable, which can’t show who is returning what and why, and how valuable it is to offer them that return for free, or which service their returned item should be shipped on, et cetera.

The third core trend is the development, adoption and integration of AI into supply chains. At Blue Yonder, we break out AI into two kinds: Predictive AI and Generative AI. Predictive AI draws on historic data to determine what’s likely to happen next, helping to forecast and plan. Generative AI is arguably the more interesting type, which has the potential to draw upon different data sources and model outcomes – so that you could ask a question like “where shall I put lockers in this city” and draw upon millions of datapoints, from traffic to weather to crime to regulations to competitor presence to generate a new answer. Or, to take the returns example: “how can I reduce returns costs by 12%?”

In that context, with interoperable systems and AI pulling in the right data, it can link to your CRM and ERP to determine which customers might be returning fraudulently, connect to your WMS and planning operation to determine which products to buy and stock fewer of, based on anticipated return rates, hook into the TMS to make sure high-value items are back in stock quickly and the rest is consolidated with maximum efficiency, and your inventory and order management to understand what’s available in store, and how that could be replenished from returns.

AI is the enabler for true orchestration, which powers the kind of efficiency that a new age of ecommerce will demand from logistics providers and retailers alike. Once data is truly shared between systems and systems are interoperable, it will be artificial intelligence that is able to connect the dots and suggest the right path forward, drawing on information that is beyond the potential of people to digest, process and calculate.

Today supply chain businesses sit on a spectrum of maturity depending on the extent of their use of interoperable technology, and AI integration. Stepping up these levels is the pathway to achieving a truly cognitive supply chain.