Time matters more than distance

Barry Grant

Posted by: Barry Grant / November 24, 2019

As today’s on-demand economy increases consumer expectations around experience and fulfillment, Order for Pickup is growing in demand for retailers, restaurants, grocers, and more. This new extension of the brand experience falls short when attempted without the knowledge of when, precisely, the customer will arrive.

Common distance-based location technology have been adapted to try to answer the question, ‘When will my customer arrive?’ But they often fall short because these tactics use distance as a proxy for actual time. When you accurately predict customer arrival time, you decrease wait time, while also increasing operational efficiency.

But the true benefit is to your brand, which, if you deliver upon the customer promise and experience that on-demand consumers expect today, drives more return visits, increased purchases, and stronger brand loyalty. Across tens of millions of orders captured by Rakuten Ready, the data very clearly shows that customers who wait less than 2 minutes to receive their order are 4 times more likely to reorder in the future than customers who wait 10 minutes.

Changing landscape

The online shopping experience has trained the consumer to have high expectations:
We order things online and it just shows up at our doorstep – no friction, right? Not always.
The on-demand economy has demonstrated retail shipping that doesn’t arrive when promised, and restaurant orders that come delivered incorrect, late or cold, which results in decreased visits and lost consumers.

According to the Rakuten Ready 2019 Time Study customers are already waiting too long. With over 20% of customers waiting at least 6 minutes to pick-up their orders, the study highlights that Retail, QSR and Grocery brands must invest now in infrastructure, marketing, training and technology to meet the incredible rise in popularity of Order for Pickup programs or risk losing to their competition. Which brings us to the ‘new challenge’ we raised above…

Without accurate customer arrival information, assumptions are made that impact customer service, operational costs, and processes. For instance:

  • Orders are fulfilled immediately as they arrive in the system, regardless of actual customer ETA; yielding soggy fries or overflow in your order holding area
  • Order packing does begin until the customer arrives, leading to long wait times.
  • Adopted technology triggers alerts to employees at the wrong time – either misfiring due to incorrect signals, or alerts store employees after the customer has arrived

Even if your organization already uses estimated arrival times to drive optimal operations, we find that most brands are still struggling. They struggle because they use geofencing combined with distance-based location technology such as cellular and beacons to predict arrivals.

While distance-based location technology have great applications in other areas, they fall short in real-time customer fulfillment, because they are not designed to answer the simple question “When exactly will my customer arrive?”

Distance is a poor proxy for time

Arrival time depends on many factors. Some are known, and based upon different routes or modes of transportation, while some are not known and can’t be predicted, like when your customer makes other stops before they come to you. To simplify this challenge, many brands have mistakenly used distance from a location, as a proxy for time.

Unfortunately, distance is a poor proxy for time. Anyone who lives in a bigger market can attest to this, because 1 mile can take 1 minute to drive, or it can take 50 excruciating minutes…

Distance-based location technology does not take speed into account. Constantly pinging the user’s device to get their trajectory will generally work, but could also create a bad customer experience by offering inaccurate arrival information and quickly draining the customer’s device battery.

At Rakuten Ready, our ARRIVE solution is the only time-based location-based technology (vs distance-based). ARRIVE doesn’t just take distance into account, but also includes elements of time and human behaviors, to predict exactly when customers will arrive.

Let’s dig into why distance-based location technology can’t help answer — “When exactly will my customer arrive?”

How distance-based location technology is being used

Distance-based location technology leverages commonly found technology such as cellular, WiFi, GPS, and beacons to create a geofence that determines where a customer is in relation to a pickup location such as a store or restaurant. A geofence is a virtual perimeter around a real-world location – such as your store, restaurant or any point of interest. You can visually see this in the diagram below.


We find that geofences don’t give organizations the level of detail they need to provide an excellent customer experience, leading to these two problems:

  • False alerts: When regions are too large, customers could be in the area defined but not headed to your store; maybe headed to the bank first – yielding false alerts to your system. This can lead to a wasted resource waiting for a customer that isn’t coming yet.
  • Customer already arrived: Customers might have arrived inside the store before any notification is sent to your system due to the region being drawn too small. This scenario can lead to a customer waiting for your teams to serve them after they arrive.


Diagram: False alerts (left), Customer already arrived (right)

Let’s briefly take a look at how the underlying technologies pinpoint your location and impact arrival times.

Your location is determined based on triangulating data from three cell phone towers. The accuracy can range depending on your location it can be as accurate as within ¾ of a square mile to much further away in a rural area.

It is great at determining a general area where someone might be located. Even emergency services know – it’s not good at pinpointing someone in need. With a wide range, it makes it hard to act on that data.

This works similar to cellular triangulation, but instead uses wireless access points and wireless hotspots to determine a location. In urban areas, WiFI can be more accurate than cell phone towers. This method is more geared towards closer ranges.

GPS (Global Positioning System) is a system of about 30 satellites orbiting our Earth. When someone mentions GPS, they typically refer to using a device or a GPS receiver to locate 4 of the satellites in space, to determine where one is located using a process called trilateration.

Due to the stationary nature of the satellites, GPS is very accurate at determining where one is located if you have direct access to the satellites (not being blocked by tall buildings, mountains, trees, etc).

Beacons take advantage of BLE (Bluetooth Low Energy), meant for short-range communications. Since Apple debuted their version of iBeacon in 2013, there are many different standards on the market.

Beacons have a shorter range and can be as accurate as 1m. They require hardware that needs to be installed throughout a store to determine location. However their shorter range creates that scenario where the customer might have already arrived. They rely on customers having Bluetooth turned on, and devices having the capability to reliably sight the beacon, which reduces their overall reliability.

You need more than just location

Though technologies such as beacons and GPS can be very accurate, they cannot predict dynamic human behavior:

  • Humans don’t travel in straight lines: It is never as simple as someone is X distance away we know it will take Y time to arrive. Additional factors can apply including traffic, accidents or many other reasons for someone to take a different path.
  • Different routes are taken: We cannot assume that people take the most direct route, in reality we have driving patterns that deviate from what could be the norm. Even if a customer follows a navigation device, it’s possible they might choose not to follow it or miss a turn.
  • Trips are combined: Order for Pickup is often combined with other errands such as picking up children from daycare or even filling up gas. This means a customer could be down the streets but minutes away due to their errand.


Diagram: Human path variation (left), Different routes (middle), Combined trips (right)

Besides dynamic human behavior, distance based technologies also cannot factor conditions on the road. It doesn’t know how to take into account congestion or even the difference with the time-of-day. At its heart, all distance-based location technologies can only tell you distance. It’s very good at answering “Where am I?” or “Where is my customer?” In order to design a proper Order for Pickup program – you need predictability as well as location accuracy.

Our Approach

Rakuten Ready is different since we fundamentally solve and optimize “When will my customer arrive”. Our machine-learning technology models routes to each of your store locations, to understand all the various ways to get there, by leveraging public information (maps, traffic, day of the week) and your customer’s location via your mobile application.

Using our sophisticated and trained engine, via your app on the customers’ device, store employees can accurately predict arrival times 99% of the time, all while ensuring near zero battery drain on mobile applications. This experience creates a win for customers who don’t want to wait for their orders. ARRIVE is available in the cloud with an SDK that makes it easy to integrate into your existing applications. No hardware needed.

If the time-based location technology isn’t enough, here are other features that are helpful to retailers, restaurants, and all customer experience focused industries:

  • Accuracy of arrival time: Nothing else on the market can predict exact customer arrival time better without having to constantly contact a customer’s mobile device for a location update.
  • Accuracy of Data: Understand dwell time or how much time customers spent within your location with the most accurate arrival times.
  • Machine Learning: Out of the box, our solution delivers an accurate model which continues to learn and improve arrival time predictions based on customer trip data.
  • Fully cloud-based: No need for hardware. No need for maintenance. It’s all available for easy integration into existing applications.

Benefits to time based location technology

Once you know when a customer will arrive, opportunities for how that metric can improve your customer experience or store operations, are endless. Here are just some of our favorite benefits to incorporating Rakuten Ready’s ARRIVE:

Meet your customer’s expectations
We live in an on-demand economy. Consumers want everything immediately. You can get an alert when your customer is on the way, enabling you to pull together the order before they arrive. You can now greet customers when they come to collect their orders. Reward your customers with time back to increase brand loyalty and profits..

Uplevel your customer experience
Arrival times can be tailored to a white glove experience. For example, apparel retailers can enable customers to reserve items online and try them in-store. Restaurants can have food hot and ready when customers arrive – rather than having them wait while their order is made or leave with an order that is cold. A hotel can prepare a room ahead of customer check-in, and ensure their favorite wine is waiting in the room.

Decrease operational costs
When you measure arrival times, wait times, and overall last mile experiences, you can leverage that data to optimize staffing, training, and processes. When you improve those three things, you can deliver optimal experiences, which leads to return customers. This means more loyal customers and less delivery commissions, which increase sales and saves your business money.

Improve your Order Management
Eliminate soggy and cold food, by timing food to be made closer to pick-up. Your order flow can turn into a ‘first-to-arrive, first-out’ model, rather than a ‘first in line, first out’ order management system. A win for brands and customers.

Ready to learn more about when your customers will arrive?