7 strategies to guarantee research impact
Originally posted on LinkedIn. Follow me to stay up-to-date with future articles.
Every 6 months many companies have performance reviews. One question that is commonly asked in the self, peer, and manager assessment is, ‘what impact did you/this person have in the last 6 months?’. It is during this time that researchers will review research reports and connect with cross-functional partners to determine the impact their research had. Tracking impact can be difficult for researchers because it is a discipline that has impact through influence, and there is often a lag between when insights are delivered and action is taken.
In this article, I outline 7 strategies to guarantee your research will have business and product impact that can be tracked. While the definition of research impact may differ between companies and industries, in this article, I define it as the ability of research to inform business or product decisions, or influence outcomes.
1. Associate research with an Objective and Key Result (OKR)
OKRs are a goal-setting framework used by individuals, teams, and organizations to define measurable goals and track their outcomes.
Associating research with an OKR ensures that the research ladders up to a goal the company cares about. This may be straightforward, but there are times when cross-functional partners and researchers lose sight of this, because of the exploratory and iterative nature of product development and research. If the researcher is unable to map the research to an OKR, the researcher should question the need to conduct the research or align with cross-functional partners on why an exception is being made, which may include conducting research to inform a future OKR.
In addition to OKRs, research might align with strategic initiatives in an organization. Regardless of the framework, research should rarely be conducted in isolation from strategic priorities, for doing so will increase the risk the research does not have an impact.
2. State the desired impact
When planning a research study in addition to outlining the goals, questions, audience, methods, etc… the researcher needs to state the desired impact of the research. More specifically, what business or product decisions will the research aim to inform, and what data collected will enable these decisions to be made? Per our definition of research impact, specifying the decisions to be made is defining the impact of the research before it is even conducted.
It can be tempting to include a generic impact statement, such as ‘This research will inform the international product roadmap for the next fiscal year.’ This is not specific enough and it is not a decision to be made. The researcher and their cross-functional partners need to work closely together to identify specific decisions, such as ‘In which country should we locate servers in Europe to best address data residency and privacy concerns of our European customers?’
This can be a challenging process and it can be easier to say, “we don’t want to introduce bias” or “let’s see what comes out of the research”. However, a decision is a determination to be made — if the determination is made before the research, that is biased; and if you do not know what you are looking for, you will not know if you have found it.
It should be noted that if the cross-functional partners are mostly confident in the decision to be made, or there is an industry standard, the researcher should question the need to conduct the research. Research should produce net new insights and not be “good to know”, “interesting”, or validate something that was already known. If research is validating something that is already known or is not acted on, it could be argued that the research did not need to happen and has not had an impact.
There may be times when research may not allow a decision to be made immediately, but will rather allow future decisions to be made. For example, user journeys or profiles may allow cross-functional partners to decide whom they are designing for. In such cases, the desired impact includes a sample of future decisions that can be made based on the research.
3. Identify a decision DRI (Directly Responsible Individual)
A decision DRI is the person who is responsible for making a decision based on the research and other inputs. In most cases, this will be either a designer or a product manager (PM) depending on the decision to be made. For example, if the decision is informing a product requirements document (PRD) the decision might be made by a PM. If the decision is related to a design decision that needs to be made, the DRI will likely be a designer. While a researcher may highlight opportunities or make recommendations, it is rarely the case that a researcher is a decision DRI.
There can only be one decision DRI for each decision to be made. There could be one or many cross-functional partners who are decision DRIs, but there can only be one per decision to be made. Others may inform the decision to be made, but there should only be one person who has to decide to ensure accountability and avoid ambiguity.
Being a decision DRI entails more than making a decision at the end of the research. Responsibilities of this role also include reviewing research materials, attending some of the research sessions, participating in analysis sessions, and workshops, as well as attending the readout. If the decision DRI is unwilling to make this commitment, the decision DRI and researcher should question the need to conduct the research. However, accommodations do need to be made for decision DRIs’ other commitments and schedules — they have their work, and as such the researcher may need to use techniques, such as daily summaries and highlights, as well as be clear about what feedback they need from the decision DRI.
4. Create an analysis plan
Before conducting the research, the researcher should create an analysis plan. An analysis plan is a document that outlines how the data from the research will be analyzed and how it maps to decisions to be made, as well as ensuring the data collected is used. For in-depth interviews, this may be a mapping of discussion guide questions to decisions. For a survey, this may be a mapping of questions, plots, statistics, and tests to decisions. This process ensures that the right data is being collected to make a decision and speeds up the process of analysis — or at least the primary focus of the analysis.
The analysis plan does not need to be a separate document, but could instead be integrated into the discussion guide, notetaking templating, or survey design, to name a few examples.
This is not to say that the research won’t produce other valuable insights that will inform business and product decisions. A researcher should explore unknowns through exploratory analysis and probing questions, and make time during the analysis phase of research to understand the implications of these learnings. Yet, it should be acknowledged that there is a risk that these learnings will not be acted on if they do not align with current business or product priorities or unless there is a reprioritization of strategic priorities based on the research — the latter of which is high-impact research.
5. Make actionable recommendations
Cross-functional partners are not merely interested in learning about their users. They want to hear what action should be taken by the researcher as a subject matter expert. As such, every insight should be accompanied by an actionable recommendation.
These recommendations do not have to be written in isolation. Hosting a workshop with cross-functional partners and decision DRIs is an opportunity to discuss insights, recommendations and the decisions to be made as a result of the research. This should be done before more broadly sharing the research with others.
6. Outline what, so what, and now what
Every research report should include a tl;dr — too long; didn’t read. The purpose of the tl;dr is not for the reader to skip the research report, but rather to direct them to the relevant content.
The tl;dr highlights what the research found. It should include an insight statement, written as a succinct expression with the syntax of “what + why” to be descriptive enough to stand alone and invite the reader to learn more (for more information on this insights taxonomy read, The Power of Insights: A behind-the-scenes look at the new insights platform at Uber). The tl;dr should come immediately after the cover or title — it should be the first content of substance the reader sees.
Immediately after the tl;dr should be a section that highlights the impact of the research, and the decisions that were made — this is the so what. In addition to highlighting the impact, the research needs to include the status and next steps — this is the now what.
It is not solely the responsibility of the researcher to write the so what; now what content — the decision DRI shares ownership. It is, for this reason, the decision DRI must be an active partner in the research because they need to be able to use the research to make decisions and defend the decisions made.
While a decision implies that some action is taken, it should be noted that deciding not to act is also a decision and as such is a plausible impact from research.
7. Take action
Nothing is someone else’s problem.
The best researchers are the ones who can put on the hat of a designer, engineer, or product manager, creating a mock, building a prototype, or writing a PRD when needed. This behavior is not limited to researchers but applies to all functions.
For example, it used to be the case that when a user opened the time input field in Airtable the default selected item would be 12:00 AM, regardless of the previous time selected. If the user wanted to change the time from 5:00 PM to 5:30 PM (the next value in the list) the user had to scroll. This was a source of customer frustration. As part of the hackathon project, I implemented and shipped a change that opened the time input field to the nearest 30-minute interval. So if 5:00 PM was previously selected, the time input field would open at 5:00 PM. If 5:05 PM had manually been inputted, the time input field would open at 5:00 PM — the nearest value in the list.
Researchers do not need how to code to take action. Researchers on my teams’ have created interactive Figma prototypes and worked with others who can code during hackathons, to name a few examples. Taking action requires going beyond making an actionable recommendation.
In summary, if you follow these 7 strategies, when it comes to writing your next performance review you will be able to speak to how your research has contributed to the outcomes of OKRs, list and link to the decisions that your research has informed, and note how you have contributed to business and product outcomes. Furthermore, your research function will be seen as critical to making business and product decisions.
Many thanks to Adrienne Michelle St. Aubin, Christopher Geison, Doruk Gurel, Grace Chang, and Peter Levin for their contributions and feedback on earlier drafts of this article.
Andrew Warr is a research executive with over 15 years of experience leading impactful, high-quality projects across multiple product lines at hyper-growth companies, and over 10 years of experience building, coaching, and managing strong-performing marketing, operations, and user research teams.
If you are looking for an insights leader, look no further. I am #openforwork. My resume can be found at www.andrewwarr.com/resume and more information can be found on my LinkedIn profile at https://www.linkedin.com/in/andrewwarr/
Did you enjoy reading this article and want to read more? Check out some of my other articles on my Medium page. All future articles will be posted as LinkedIn notes. Follow me to stay updated.
