3 Ways to Make Your Research Better Today

In my new role at HubSpot’s User Experience (UX) Research team, I was recently asked to share how we approached research at McKinsey and during my three startup stints at Canary, Catalant and Happie.

While I never had the formal title of “Researcher,” the majority of projects during my time at McKinsey advising Fortune 500 companies, as well as my work as a PM / PMM involved the use of qualitative and quantitative research to develop recommendations. I was pleasantly surprised to find my team greatly enjoyed my presentation, so I’ve decided to share it here!

Here are my learnings on why research is important and how to approach research effectively.

Organizations are very, very hungry for research. Why?

Research can help answer questions that help organizations create value against a finite amount of resources. Here are the three broad buckets of questions that people typically seek answers to:

Without research, organizations would be left shooting from the hip into the dark.

While research is valuable, a lot of factors can hold researchers and research projects back. To list a few: volume, lack of time and resources, breadth and complexity of the problem and solution space, and challenges in team or stakeholder management. What we want to learn may completely overwhelm the resources that are available. How can we overcome these challenges? Let me take a stab at a few suggestions.


#1. Use a structured approach to problem solving

“Structure” can probably solve half the problems I mentioned above. Structure helps teams collaborate effectively by bringing clarity and prioritization on the following dimensions:

  • Question: What is the question you are trying to answer?
  • Issues: What are the main issues / hypothesis that are relevant to focus on to answer the question?
  • Analysis: What is the analysis that should be conducted in order to identify approaches to address the issues / hypothesis?

There are two useful frameworks that are incredibly useful to bring structure in a research context:

Seven Steps to Problem Solving

This McKinsey framework is a step-by-step problem-solving approach that applies logic rigorously to get from question to actionable recommendations. It is a framework I know by heart. If you looked into my DNA… it might be there. Not only do all consultants at McKinsey learn in their first week of training, but teams continue to refer to it in each client project. I hope to cover it in more detail another time. For now, you can learn how to go through the Seven Steps in this McKinsey document.

The Seven Steps of Problem Solving

Issue Trees

An issue tree is an approach to deconstruct a question into discrete addressable components. Below is an example. You might ask — When do you create an Issue Tree? Typically in Step #2 in the Seven Steps mentioned above. If you’re not using the Seven Steps, get in the habit of creating an Issue Tree any time you are trying to tackle a big question. You could go through the exercise on your own, or with a team. In my early days at HubSpot, I went through several iterations of Issue Trees to try to wrap my head around big hairy challenges the Growth Team is trying to tackle.

Don’t feel bound by the framework. Find a framework that works for you. If you’re interested in trying out building one, remember to be MECE (Mutually Exclusive, Collectively Exhaustive) and follow the 80/20 rule. Read more about MECE and 80/20 here.

How powerful would it be to be able to: …tell people where you are in a problem solving process? … tell people your plan of attack? …tell people why you aren’t addressing a question they want to have answered (push back)?


#2. Integrate mixed methods and sources to maximize impact

This is what I mean by “methods” and “sources”:

  • Methods: Qualitative (Ex. interviews, focus groups, ethnography), Quantitative (Ex. surveys, internal data, external reports)
  • Sources: Primary data (Ex. observational, interviews, surveys, ethnography, etc.), Secondary data (Ex. market reports, expert interviews)

In all the projects I recall — whether at McKinsey or a startup — we used a multi-pronged approach to figure out an answer to a question. Here’s why:

Progressive learning

Explore different approaches (methods and sources) depending on where you are in the problem solving process, as well as the accuracy and granularity of information you are looking for. For example, if you are looking to develop initial hypothesis on how to improve user engagement, it could be more appropriate to conduct a couple user interviews with a small sample before proceeding to do a fully blown-out survey or experiment. Carefully structured and planned out research can facilitate targeted learning along the way.

Cross-validation

While properly designed research should stand on its own, the findings can be even more robust if they are compared against other sources. At McKinsey, we constantly cross-validated research against other sources, internal or external. While I appreciate the “scrappier” approaches I have seen in startup contexts, and recognize that validation can also occur in the form of rapid iterative experimentation, I do believe that it is worth the investment to cross-validate upfront, particularly depending on the size of the investment or how “big” the bet is.

Ideation

Ideation is a process of convergence and divergence. While we often think of “Data-driven” as an approach to find “right” answers (convergence), I think aggregation of a variety of data can also help trigger inspiration (divergence) to lead us to more creative hypothesis and solutions. For example, when I was working at Canary to identify opportunities in International Markets, I brought folks from across the organization together to go on a “Gallery Walk” to review market and tech trends, our progress in retail and business development (BD) strategy, customer experience and product usage data. The results? We came up with new ideas to improve our user experience and go-to-market strategy due to the exposure to a plethora of ideas.

Stakeholder alignment

One of the most important reasons for bringing together a variety of perspectives is to facilitate understanding among a diverse set of stakeholders. In one of my most memorable projects at McKinsey working with a consumer goods company, we set up an ideation workshop, bringing together data from multiple “lens” — Consumer (user), Market, Technology. The participants included folks across the organization — Product, Marketing, Engineering, Operations and Finance. It was incredibly powerful to see how data could establish a common foundation among people with different perspectives, and help the group collectively identify issues and opportunities together.


#3. Develop clear recommendations and drive implementation

Don’t let everything (insights, $, pride…) go down the drain. No matter how brilliant the research and findings are, without delivering results, none of it matters.

Here are a few tips on developing recommendations and leading implementation:

Communicate using the Pyramid Principle

The Pyramid Principle is an incredibly useful framework for effective visual and verbal communication. It’s all about top-down messaging. Start with the conclusion first, then make sure the building blocks are there to support the overarching message. A simple way to do this is by making sure the hierarchy is maintained in the structure of your powerpoint slides or written documents visually. In the case of powerpoint, apply the rule of thumb “one-slide, one message”, to make sure everything builds up to the message in the slide heading. If you’ve structured your documents following the Pyramid Principle, the delivery should be straightforward. All you have to do is read from top-to-bottom, from one-slide to the next.

Draw actionable implications

There are too many instances where I have personally found myself excited about the synthesis of research that is filled with really interesting facts that satisfy my curiosity. But “so what”? What does this mean for the team or organization? Go the extra mile to draw implications that are actionable for the team and can lead to impact.

Develop an example roadmap

So you’ve drawn the implications. But how will the team go about tackling the recommended path? If you think it’d be helpful to get the ball rolling, put together a straw-man roadmap that outlines the main activities, milestones, and owners and stakeholders. It’s not easy for a team that is used to going through the same motions to come up with a plan to roll-out a new initiative.

Involve & get buy-in from the right stakeholders

Hopefully, you’ve been involving the relevant stakeholders through the entire research process (or each of the Seven Steps if that’s the framework you decided to use). It’s important to keep them involved along early parts of the journey to make sure you are focusing on the right questions, but perhaps even more so in the latter parts if you hope the team will act on your recommendations. The most successful projects I recall from my time at McKinsey were those where we made sure our clients were onboard every step of the way, and not just for the presentations. On the other hand, I’m unfortunately not sure if some of the research I did at Catalant went anywhere. This is partly due to the short duration I was at Catalant, but also more likely since I was not involving the right stakeholders.

Conviction

Lastly, this is more of a tip on how to deliver your findings and recommendations. While researchers can lean on data to substantiate their claims, I also think that conviction is a very important attribute for leaders who want to drive change. The fear associated with the uncertainty can be a key barrier to for teams who consider change. However, leaders with conviction can help their teams overcome uncertainty and enable them to focus on implementing the best course of action. During my time at Happie, I recommended changes to the way the tech-enabled service operated and the design of the internal-facing product. When I made the recommendations, I not only packaged the data to support my claims, but tried to make sure my message to the executive team was delivered with conviction.

Go the extra mile to develop actionable recommendations and lay the ground for your insights to make a true impact.


Quick Recap

Research helps organizations solve for impact
Researchers bring an external perspective and drive change

#1 — Use a structured approach to problem solving
#2 — Integrate mixed methods and sources to maximize impact
#3 — Have clear recommendations and drive the implementation of them

Please add your thoughts or questions below, and I’d be happy to follow up with them!