Whether you’re a global enterprise or a small business, it’s vital that you understand your company’s total potential and what opportunities are most likely to get you there. Real estate isn’t just one of your largest operating expenses. It’s also one of the biggest factors in your success or failure.
Location is everything. And effective market planning and site selection is key to investing in the right location. While these two processes aren’t always handled by the same teams or individuals, they go hand-in-hand. Whatever your role, Tango Predictive Analytics helps you from start to finish, equipping you to analyze market-level opportunities and zero in on specific sites within each market.
In this guide, we’ll explore:
- Market planning: optimizing your potential in new and current markets
- Site modeling: creating accurate comparisons between stores
- Sales forecasting: measuring specific opportunities
- Site selection: prioritizing the best locations for your business
- Site selection software: supporting decisions with data
Let’s start by digging into market planning.
Market planning: analyzing broad growth opportunities
Market planning is the process of finding real estate opportunities within a trade area, city, state, region, or country. This involves a combination of assessing current stores and exploring new ones. Depending on the maturity of your business, market planning may be a routine process you go through periodically before beginning site selection, or it may be part of a case to secure funding, win grants, or make other business deals.
Retailers use a variety of approaches to identify markets that can support their stores. You might use void analysis to estimate how much room a market has for more stores like yours. Or you might use a form of white space analysis to generate a count of the markets you could grow into. Sometimes the market planning process begins with a predetermined list of cities or markets you want to vet. Other times it starts with basic filters to generate a list. You might also examine underperforming stores and look for better trade areas nearby or consider whether a remodel would help meet your goals for the market. Whatever your method, your goal is to eliminate irrelevant markets, leaving only the opportunities that appear most viable on paper.
Think of it like trying to fill a position in your company. Without any qualifications, there are millions of potential candidates. But you don’t want just any candidate. You need to narrow the scope of your search before you take the time and energy to interview individual people for the role. You don’t want to miss out on potential stars other companies overlooked, but you also want to ensure your pool of candidates are a good fit for the role and your company’s specific needs. So you filter candidates with the ideal qualifications, and depending on the number of candidates you wind up with (and how strong they are), you may adjust these filters throughout the process. Bad filters give you bad candidates. And if your hiring process has issues, you may wind up hiring people who aren’t a good fit (and you might not realize that until months or years later).
The trick with market planning is that you need access to reliable datasets in order to filter cities, trade areas, and markets by the right criteria—namely, demographic data and spending data. You also need a rough model of what a good market looks like. If you were hiring for a position, you might model your ideal candidate around your best employees. In market planning, you’re looking for markets that can best support a store, so it makes sense to consider what markets are working for your stores right now. What do the demographics and spending habits look like in the markets you’re having the most success in right now?
This is why market planning works best when you use software like Tango Predictive Analytics. Software can combine the datasets you need with GIS mapping, empowering you to visually explore markets. You can see where the demographics you need are concentrated, the actual locations of potential competitors, and the market’s total geographic area.
It might be nice to imagine you only need to conduct market planning every few years, but markets can change fast. A city’s largest employer could change ownership and lay everyone off, causing a rapid exodus of your target demographic from the area. A large new development could go in, enabling a city you’d written off last year to support your business for years to come. Maybe a new city joined the list of fastest-growing retail markets in the US, making it a prime candidate. As these changes occur throughout cities around the world, it impacts which markets are most worth investing in.
Market planning results in a list of markets worth exploring. Then, it’s time to build a model that lets you analyze specific sites—so you can determine if a trade area can actually support your business, and what performance you could expect at a particular location.
Site modeling: creating accurate comparisons between stores
Site modeling is one of the most critical aspects of market planning and site selection. It’s also one of the steps retailers are most likely to get wrong. A bad site model leads to inaccurate sales forecasts, and ultimately, costly real estate decisions. When your projected performance doesn’t match actual performance, your site model is usually to blame.
The point of a site model is to identify the factors that impact a store’s performance, and then accurately weight their impact based on your current portfolio—all so you can compare potential locations to your model and forecast sales.
At Tango, we generally group these factors into two distinct categories: site characteristics and trade area characteristics. A store with the ideal physical and operational qualities in a bad trade area is unlikely to perform well, and the same goes for a store in a good trade area with poor physical and operational qualities. It doesn’t matter how big the parking lot is if there’s no one around to visit your store. And if you’re surrounded by people, but they can’t see or access your store, that’s going to hurt you, too.
We use a series of about 10–30 questions to analyze the most relevant site characteristics for each store in a customer’s portfolio. Then, analyzing current store performance, we weigh these characteristics based on their impact on performance. Leveraging our datasets, we do the same for the trade area surrounding each store. This lets us give each store a score for site-level and trade area characteristics and compare similarities and differences between them.
Site modeling becomes exponentially easier with machine learning. In fact, given the volume of data to process between site and trade area characteristics, it would be impossible to build and utilize a reliable site model without training a machine learning algorithm. Using the weighted criteria, your model should accurately “predict” the actual performance of a sampling of your stores.
Once you’re confident you have a trustworthy model, you can use your model to forecast sales at potential locations.
Sales forecasting is the part of the process that enables you to validate or rule out particular sites within your target market. You’ve already determined that the market should be able to support one or more of your stores—now you just need to identify where your stores should go. Which location puts you in the best position to access demand? Does a viable location actually exist within this market? Sales forecasting is the only way to find out before you decide.
Suppose there are a handful of potential sites within a trade area. Even if the facilities have the exact same physical specifications, their performance will vary based on how various demographics move throughout the trade area. You can’t simply find a location next to demand. You have to set up shop at a site where people can regularly encounter and access your business.
This is why mobile movement data is a vital component of sales forecasting. By tracking anonymized movement data, you can see the concentration of specific demographics over time, even filtering by day, time, or season, giving you a fuller picture of the actual size of the opportunity surrounding a potential site. Rather than basing your calculations on high-level demographic information, mobile movement data leads you to the right corner of a busy intersection or the ideal side of the highway.
All of this information needs to be fed into your site model so the algorithm can yield sales forecasts for the particular locations you’re analyzing. And as long as you haven’t overtrained your model or overlooked any key site selection criteria, you can lean on your forecasts to make reliable decisions about where to open, merge, relocate, or close your stores.
Site selection lives at the intersection of your location strategy and sales forecasts. You’re not just looking for one new location—you’re building a prioritized list of the next big real estate moves your business should make. You want to make the right moves in the right sequence. And that will require you to balance business goals (such as expanding into a new state or country) with your capabilities and limitations (such as the added strain on your supply chain) and each location’s projected performance.
While market planning generally provides long-term guidance, site selection is a process that retailers must frequently revisit. You don’t ever want to be in a position where you’re not ready to grow, and large retailers will always have numerous projects at various stages of development. Regularly conducting site selection ensures that you have an approved, validated list of locations “in the hopper” at all times.
Note: You’ll also want to estimate how moving forward with a specific location will impact any nearby stores you already have. Will the new store’s performance outweigh the cost of any cannibalization? When calculating cannibalization (including sister store cannibalization), you’ll want to consider each store’s actual serviceable area and avoid using a cookie-cutter approach.
Years ago, site selection relied more on instincts than algorithms. The pros “just knew” where to expand. And some of them are still pretty good at it. But there are a couple major problems with this:
- These are multi-million dollar decisions. Being wrong comes at a huge expense.
- What happens when they leave the company? How will you replicate (or even learn from) their success?
Using site selection software gives you projections you can justify and results you (or anyone else) can repeat. Every vendor and system works differently, but with Tango Predictive Analytics, our experts work with yours to build and train unique machine learning algorithms that fit your business.
The results of our sales forecasts may sometimes surprise you, but they’ll never leave you confused: we take time to analyze and articulate differences between your expectations, our projections, and actual performance, zeroing in on every variable that counts and ruling out those that don’t. Our calculations come from tangible, explainable figures—so your decisions can be rooted in them, too.
When people responsible for market planning or site selection move on, their successors can easily pick up where they left off. Everything they need to understand and validate old decisions stays in the platform, so your business can continue to grow intelligently and efficiently.
There are a lot of site selection vendors available. Some take a more traditional approach to market planning and sales forecasting. Others use cutting-edge tools and processes to deliver results fast. So how do you know which solution is best for your business?
For the most part, choosing site selection software should come down to two things: can you trust its results, and can you explain them? Here are the key factors to compare.
Site selection software is only as reliable as its underlying datasets. You need to be able to trust that you’re working from the best information, or else your models and forecasts will always be wrong. Some tools lean more heavily on census data. Others use datasets that update more frequently. The types of datasets you have access to can completely change the kinds of insights your software can reveal.
At Tango, we use a combination of datasets including:
- Consumer spending data from Experian
- Mobile movement data from Near Intelligence
- Retail location and co-tenant data from ChainXY
- Traffic flow data from INRIX
- Lifestyle and economic indicators from Synergos Technologies
- Purchase data from Yodlee
- Canadian market data from Applied Geographic Solutions
- Shopping center ratings from Green Street Advisors
- Census data
- Your own customer data sets
We also use a custom survey to learn about each site and assign it a score based on its ability to access the demand in a trade area. This is unique data that can only come from your site analysis, and it has a major impact on the reliability of your results.
Your software’s accuracy doesn’t just depend on the data it uses. It also depends on how it uses that data. Many vendors will apply a single algorithm to all of your locations, reusing that same algorithm across every type of business they work with. This significantly lowers the accuracy of their results because it fails to understand how certain variables change in relevance based on facility characteristics and industry.
Tango Predictive Analytics uses a series of algorithms that build on each other, mixing and matching from more than 100 options to find the combination that best fits your sites and your business. This process dramatically reduces the margin of error by ensuring each variable is properly weighted, producing stunningly accurate sales forecasts.
Site selection software should help you justify decisions and strategically prioritize deals. But with what’s at stake, you don’t want to be stuck saying, “We should do it this way because the model says so.” You don’t have to manually perform calculations or apply algorithms, but you should be able to understand how those algorithms are weighting variables and why that approach is best. And you’re not going to get that with every vendor.
Some traditional solutions are great at explaining how results came to be. It doesn’t mean the results are accurate, but you at least get helpful context. A few modern solutions essentially have a “black box” process you don’t get any visibility into, which leaves you with numbers you can’t explain.
With Tango Predictive Analytics, our experts work right alongside you, helping you craft and understand the ideal model for your business. We don’t accept results we can’t explain, and you shouldn’t, either. We also offer our customers a free Portfolio Review service, which they use to analyze any discrepancies between projected and actual results, diagnose anomalies, and fine-tune and refit their models.
Building quality site models takes time. But the software you choose will determine how much time—and that directly impacts how easily and frequently you can refit or recalibrate your models. When market conditions change, more competitors enter a space you’ve been eyeing, or you learn new information about your customers, you may want to quickly pivot your plans.
Some vendors charge high rates for even minor refits because it involves a lot more work (and time) on their end. At Tango, minor refits come with minor costs, and you’ll usually just need to conduct more involved calibrations every 2–3 years. While we take a very hands-on approach to site modeling, our reliance on machine learning gives you results faster than most competitors.
Once you’ve lined up a series of real estate decisions, it takes a lot of work to execute them. Many retailers wind up using offline, piecemeal solutions for tracking their transactions through various stages of approval and development. But there are a lot of moving pieces to coordinate, and numerous people involved—which makes it easy for mistakes to slip through.
While you can always repurpose other project management solutions, Tango Transactions is a specialized real estate transaction management software for retailers that integrates with your site selection tools. It serves as a single source of truth for everyone involved in the processes of opening, closing, relocating, and consolidating store locations. Many of our Predictive Analytics customers opt to add Transactions to manage the entire store lifecycle in one platform.
Opening, closing, relocating, or consolidating stores can easily involve millions of dollars. These aren’t decisions you can afford to make lightly. Every wrong decision is costly. And that’s why more retailers are turning to advanced site selection software like Tango Predictive Analytics. With Tango, you get the lowest margins of error in the industry, and you’ll always have the context you need to explain your decisions with confidence.
Want to see what Tango Predictive Analytics can do for you?