Research & Development: What Would You do With 80% More Budget?

Research & Development: What Would You do With 80% More Budget?

When a company switches to Shopify Plus, they typically find a great deal of their platform operation budget has freed up and are left wondering what to do with the extra savings. 

After all, several reports agree that 80% of I.T budgets are spent maintaining legacy technology stacks, so what happens when those costs are reduced significantly because they're standard features?

Where does that money go?

  • Advertising?
  • Marketing?
  • Better packaging?

These are all viable options — options plenty of Shopify Plus merchants like Oreo and Runkeeper have taken advantage of but what if you put that extra cash into research and development so you can experiment with those ideas you’ve been putting off for forever?

Once you’ve reduced a significant portion of your operating costs, you’re free to make the money you spend in other areas (marketing, product management, staff) more effective, via research and development.

So where’s an ecommerce business to start?

Research: Making Sense of the Data (Without Being a Data Scientist)

If you’ve been in business long enough to be successful, you have tons of data on your customers and their orders.

The problem is making sense of all of that data — connecting not just A to B, but B to Z and Z to H, and then pulling actionable insights out of those connections.

This is where Jonathon Morgan, co-founder of Popily and co-host of Partially Derivative (a podcast about data science - highly recommended!) comes in with insights on businesses can make more sense of their data.

Start With a Question

The biggest problem that most businesses have when it comes to data, is that they start with the data and not with a question.

Humans naturally want to see patterns in everything. If you give us a group of data points, we’ll create pictures and patterns and stories. Just look at the night sky for an example.

The problem is that the patterns we see aren’t always there, or that those patterns aren’t particularly useful. “People make a lot of assumptions about what’s in the data - you assume your data can only tell you a very specific story, and of course, if you look for it, you’re going to find that story. You might bury the lede, or find X but miss Y.”

It’s tempting to just hand over all of your data to someone and say “Here’s the data, tell me what it says.” That’s not usually a recipe for success - there’s all sorts of things in the data. Instead, Jonathan suggests starting with a question about what you want to improve about your business or product.

In other words, instead of looking for patterns in the data, you need to figure out what your current research priorities are, and start there. 

Start with a question you need answered or a hypothesis you want to test (“customers who buy product X tend to have the biggest LTV, because they’re this type of person with these needs”). You’ll get more useful insights from your data.

When Do You Get Help With Your Data?

Jonathan says the time to get help is when you’re committed to making objective decisions about the direction of your company and products. You don’t have to wait until you’re making million-dollar decisions.

The great thing is, where your options might have been limited previously depending on your funds (hiring a full-time data scientist or team, or buying expensive software, can really add up).

“Now, we’re in a team where small consultancies, technology, and apps make it much easier to go over data every day. We’re moving towards every day data-driven habits, not just changing direction once every twelve months based on a new report.”

As far as the technology, Jonathan’s company Popily, helps you make sense of your data. Keen is another data visualization option. And on the tracking side, Custora and RJMetrics can help you create a strong data infrastructure.  

What to Look For In a Data Person/Agency

If you’re not familiar with data science, you might not have any idea what to look for when you seek outside help, or what your first project should be when you start working with someone on your company data.

First off, whoever you’re working with should make the data easier for everyone to understand. People in one department might see something people in another department would miss. The more people understand what they're seeing, the better chance you have of getting actionable insights from it.

Jonathon says the first thing you should work on is data exploration.

“It should be the first project every time, and it’s important that everyone who has an interest in the outcome of that research is involved in the process.”

One way to include people, even those without technical backgrounds, in the exploration process is visualizing the data. Even if you don’t go that route, the data team should make every effort to pull outside people in and have conversations around what the data means, what might matter, what might not and so on.

“Skipping that step is a recipe for burning hundreds of thousands of dollars,” notes Jonathon, adding that “it’s unrealistic to expect one person to interpret the data and understand enough about the business to figure out exactly what about that research is going to be the most valuable to the business.”

How to Incorporate Insights Into Development and Design

You’ve got plenty of actionable insights from your research — what you need to do now is put them into action.

Approaching development and design using your data insights will help you make the most of that research (and save you money). But what does that look like in practice, and how do you know when to start incorporate it into your business?

Katie, partner at Knurture (a boutique UX/UI agency based in San Francisco), shared some tips:

"The most common mistake that I’ve seen business owners make is blindly trusting their gut instinct. By not putting your ideas in front of customers, you could waste a lot of time and money. A rigorous user testing process can take as little as a few days and could save your company tens of thousands of dollars, along with all the time you would have taken to develop your initial idea." 

But What Does User Research Actually Consist Of?

Many user research principles are discussed as they relate to apps and websites, but they can still be used for physical products — and, of course, improving your website, which is vital for keeping sales strong.

A basic user research process looks like this:

  • Often, but not always, a prototype of the product (or site, app, etc.) to be tested is created. Katie and the Knurture team strongly prefer to test with prototypes over running tests with wireframes or sketches, finding that it results in more usable, specific, and accurate user feedback. If you want to create your own clickable prototypes, InVision or similar apps can help.
  • Users are observed while interacting with the prototype. Often, they’re asked questions as they do so, or asked to narrate the process of what they’re doing and why.
  • Users are also asked questions after the test is over. Open-ended questions are the goal here — for example, “Why did you do X?” rather than “Did you do X because of ABC?”
  • Using the feedback from this round of testing, another prototype is created, and the process starts over again.

(For more on user research, check out this primer or this Note to Self episode where the podcast host participates in user research for Etsy.)

And yes, even though it’s research, it’s still part of the design and development stage, because the research is being done with hands-on prototypes — ideally, created with the insights you squeezed from your data.

Where data science gives you quantitative data, user research gives you qualitative data, which can help you get to the root of why people are behaving the way they are, rather than just telling you what percentage of people are behaving that way. 

When to Start and What to Watch For

Katie says you can add research into the design and development process at any point, but earlier is better if your product depends on UX. In the ecommerce world, this could be a product that you’re creating completely from scratch, like Muse headbands (a Shopify customer that launched on IndieGoGo).

If you’re selling something that people are immediately familiar with, then testing your product (if necessary — there’s not a whole lot to test with a t-shirt, for example) or website can wait until after you’ve achieved product-market fit. (For more about what product-market fit is and how to achieve it, read our article on the real reason people aren’t buying from you.)

Katie also notes that it’s important to watch for signs of a less-than-optimal user experience. For an ecommerce store, things to watch for include:

  • Customers buying one product from you and never returning. Unless you’re the rare retailer that sells something people need once in a lifetime, you want to create repeat customers. If they aren’t returning, it could mean that there’s a part of the process your company can make easier (or that you can improve your product).
  • You’re getting a lot of post-purchase questions about how to use your product. If this happens, it’s an opportunity to either refine your product so that it’s easier to use, make your how-to documentation easier to find on your website, or add how-to materials into your product packaging.
  • Looking at your number of repeat customers (and the feedback you get from customers) can give you an idea of how people are engaging with your product. Another idea is to send out a post-purchase survey 1-3 months after purchase and ask if they’re still using your product, how often they use it, etc. The insights you get can be used to make your product and marketing stronger.
  • If you offer a free sample, keep an eye on your conversion rate from free sample customers to paying customers. This is especially relevant in subscription-based ecommerce businesses. One example is ThisIsL, which offers a subscription service (in addition to one-time purchases), and has a free sample offer where the customer only pays $4.95 for shipping and handling. There’s no set “industry standard” for this conversion rate. For SaaS companies, it ranges from 11-50%, although it’s worth noting that there’s a large jump in paying conversions when you require a credit card for the free trial — which, obviously, if you’re charging for shipping, you’ll have to. Either way, measuring it and running experiments to improve it are a must if you offer free samples.

All of these are in addition to typical website UX problems. Signs of a bad user experience for your store would include symptoms like a high bounce rate, people are abandoning their carts, or customers have to get in touch with a support person to complete their order. And for a list of places to start refining your post-checkout customer experience, check out our article on designing the new customer experience.

How Much Should You be Spending on R&D?

It varies, depending on industry, but to give an idea, here’s the average spending (as a percentage of income) from businesses in the leisure goods industry, as collected by the European Commission:

Image via Statista

As far as some other industries go, electronics and electrical appliances is a little lower (hovering around 4-5%), and technology hardware and equipment has been at about 8% for several years now.

If you aren’t allocating money to R&D (or you can't),, you're probably losing money.

Check out these examples:

  • Credit Suisse implemented a user-centered strategy for their consumer facing website and had, among other results, a 15% increase in upsells (and a 30% increase in usage).
  • A 2009 report from SDS consulting found that when organizations invest in UX early on, they reduce the length of product development cycles by 33-50%, letting the product get to market faster (and that same report also found that getting the product to market slower means less profit, with a one-quarter delay turning into a 50% loss of profit).
  • The team of EMC data scientist Pedro Desouza developed an algorithm that helped a company predict and prevent account closures by 30%.
  • Kroger used big data to get more actionable insights from its customer loyalty program, helping them — this program has delivered over $12 billion in incremental revenue since 2005.

This is what you can do with all the money you'll save from being on a world-class ecommerce system at a fraction of the cost, but it's hard to make room for R&D when your site costs an arm and a leg to maintain.

If you’re ready to step up your game (and free up some money in the process), head on over to the Shopify Plus page and learn more today.

 


About the Author

Michelle Nickolaisen is a freelance writer and business owner based in Austin, TX.