How to do market research for a startup (from a Saas product marketer)

market research questions

Market research helps companies get out of their own heads and gain valuable insights into how their customers really think.

For new companies, market research can help determine the size of the market, the potential for market growth, and customer needs. For established companies, market research can help find new markets, determine positioning, and even what color to make a label.

If done well, market research can make yours and your company's odds of success go from a shade above zero to sky high. If you collect data without figuring out the questions you need answered -- therefore wasting lots of moolah and precious time -- research can hurt you.

These are my notes as I learn about market research

If you've already seen my earlier post on market orientation, you may know that I'm taking the Mini MBA with Mark Ritson. These posts are my way of thinking through what I've learned. I hope they're helpful to you too.

If you disagree with anything here, please tell me. If you have questions, please ask them.

Market research key lessons

The best way to test market assumptions is market research. Conducting market research and analyzing the data can provide much needed insights into what customers value and need. This increases the chances that a new business will succeed and increases the odds that changes to an established business have a positive impact.

For new businesses, the key questions you should try to find include:

  • Is my business concept viable?
  • What risks should I be aware of and what knowledge do I need to help minimize those risks?
  • Is my strategy and business model likely to succeed?

There is a process for conducting thorough market research. For new businesses in particular, market research should follow a structure that delivers the biggest impact with the resources (time and money and people) available. More on this below.

Focus groups are still relevant if done well. Focus groups can be a cheap and fast way to learn about your customers and your market. The timing of your focus groups will determine how valuable they are. When used in the middle of your research process, when you have customer segments defined and some tactical ideas to test, you can glean invaluable insights.

Good market research combines Qualitative and Quantitative methods. Qualitative methods of conducting research, like expert interviews and ethnography (getting out in the field to see how customers actually live with products), help identify information needs and business insights. The people around us have a lot of knowledge and experience that can help - why not ask? Qualitative data gathering helps shape the quantitative data needs.

Quantitative research, like large-scale surveys, provide hard data to help you make decisions. Income and revenue are examples of quantitative data, but this kind of research can also include data pick up directly from users. If you run a Saas business, you likely have tons of user data you can slice and dice and use in all sorts of fun and high-impact ways.

The process of market research

This outline is based on a Note on Market Research from Stanford University, but includes ideas from other readings, too.

1 - Prioritize information needs and determine how you'll use the data

Professor Alan Andreason of Georgetown University thinks researchers should start with the end in mind. He outlines why in his article called Backward Market Research.

Here's the main idea: Before starting to gather information, you first need to know how you'll use the data you collect and why you need to collect it. Andreason even recommends imagining what your final report looks like because knowing this helps you make collect the right data for you to make your decision.

Here's what backward market research looks like:

What decision are you trying to make? Lay out the action alternatives you might take after the study. Put another way, you probably know roughly what your options are, so write them down. I mean, you're not going to switch from building a Saas product to making essential oils.

This is not to say that you predetermine the outcome of the research. This is to clearly state your reason for conducting the research in the first place.

Figure out how the final report should present your data so you can make a decision. To do this, first write down all the people who will see this report. Then figure out what they do and don't like to see or how they process information. This includes bosses and CMOs and CEOs and CFOs and crusty old board members. If they like charts, give them charts. If they like spreadsheets, give them spreadsheets. Let them eat cake and make it easy on yourself by dishing it up to key stakeholders how they like it.

Your report outline will have holes in it that only your research can fill. Specify the analysis you need to conduct to "fill in the blanks" of your report.

Lastly, figure out what kind of data you need to gather to complete the report and conduct your analyses. Let this guide you as you build your research approach and the questions you'll ask.

Once you know how you'll use the data (i.e.: what the report will look like, which decisions you're deciding between, who likes pie charts and who doesn't, etc.), you can determine what sample data you need and you can build the analytical tools to suit. And now that you know what data you need and how you'll process it, you can build your data collection front end to feed into your process.

I like to imagine this process like a 1970s mainframe computer. Someone had to figure out how to build those city-block sized machines, but I doubt the process started with some dude pacing around around an empty room with a punch card in his hand thinking "Now what does a machine need to look like to read this?" Nah. I reckon they conceived a machine and understood what they wanted it to do for them and then they figured out how to provide the data: punch cards. Building your report first is like building your supercomputer before learning how to feed it.

Now that your report is setup and ready to rock you can move on to conducting research. Here are some tools for doing that.

2 - Conduct expert interviews

Speaking with experts and people with deep experience in a subject helps identify your own information needs and can verify your approach. Experts can often give you market, customer, and business insights that could take ages to uncover through other research methods. Who are the experts in your field or market?

3 - Develop your primary and secondary research plan

Mixing secondary research with primary research is a must, as we'll cover shortly, and this is the step when you determine how you're going to collect data from both types of sources.

But WTF is secondary and primary research? The Purdue Writing Lab defines primary research as "any type of research that you collect yourself. Examples include surveys, interviews, observations, and ethnographic research."

Secondary research gathers existing information from available sources like the internet, existing market research from other companies, data from your internal tools, info from industry bodies, government agencies, and libraries.

Be explicit with your plans. List the people you need to speak with and figure out how you'll line up those conversations. List the secondary data sources you'll start with (secondary research often leads to additional sources through citations, links, etc.).

4 - Conduct secondary research and start your primary research

Secondary research is data developed by other people before you. Look for sources including market reports, news stories, trade journals, competitor content, industry organizations, and well-known blogs.

Be careful here. Secondary research may not fit your needs and researchers may have segmented their data in ways that don't match your needs. The data may be old or it may be biased and its reliability unknown.

Despite the risks, secondary research is important. It helps highlight industry trends, market size, the competitive landscape, and it can identify where to focus your primary research. It can also help avoid gathering data someone else has already collected.

Start your primary research, as part of this step, by interviewing or engaging directly with your customers. This should deepen your understanding of the market and its needs, and it may confirm what you've learned from your secondary research.

I like to also list any assumptions I have before I start, so I can come back to them later and confirm or kill them.

5 - Revise assumptions and update the research plan

With new data in hand, what has changed? Are your assumptions correct or plain wrong?

It's OK to get your assumptions wrong, by the way. Proving assumptions wrong means you're learning and that's exactly what you want from good research.

Perhaps the secondary data you need isn't there or the people you thought you needed to speak with aren't the right fit. Armed with secondary data and first round primary data, you can correct course and make the rest of your research much more valuable.

6 - Conduct more interviews and review secondary data

This is when your research gets cranking. You've set the stage by planning, doing initial research, and confirming that you're on the right path. Now the fun part is ahead of you.

Expert interviews, customer interviews, focus groups, surveys, choice modeling, and concept testing are your tools in this section.

Continuously check that the data you've gathered meets your needs. You already know what your report looks like and what kind of decisions you need to make. Does the data collected allow you to make those decisions?

As you learn more, your understanding of the market grows and you're better equipped to check assumptions and figure out where to find more helpful data. Of course, you could go around in an endless loop of revising and repeating this process, so at some point you need to complete your planned report and make your decision.

Perfection is not attainable, but if we chase perfection we can catch excellence. Vince Lombardi

The point is not to create the perfect report with all possible data. The point is to gather enough smart data to make smart decisions that allow you to move.

7 - Review your data, check assumptions, fill in your report, make your call

If you used the backwards approach, this should be straightforward. Using the data you've collected, write your report and run your analysis. If you designed the research project well, you can be confident of the outcome.

Further reading on doing your research

Entire books have been written on how to conduct market research, so this little post is nothing more than a primer. If you'd like to learn more, here are some resources I found helpful:

This is the second post in a series on marketing. Here's a full list of the other posts. I'll update each one with links as soon as they're published.

Peter Preston

Peter Preston

I'm a Saas marketing manager at ThinkTilt, makers of ProForma for Jira. I'm also the founder of Dear Video, a recovering podcast host, and learning how to grown a brand on YouTube.
Brisbane, QLD