Research ... revisited!
The research industry needs an overhaul.
Both internal research teams and research firms tend to have a dated understanding of what research is. Quite often they get a brief (often to fill in data gaps), use their preferred methodology to execute the research (desk search for some, customer surveys for others) and present the results in an often very beautiful way but without going as far as deriving insights.
However this way of doing research is already antiquated, stuck in a time when the Internet was not yet a ubiquitous resource and when gathering data via primary research or leveraging paid sources (books, subscriptions) was perceived as high value.
Today, data has become a commodity. Simply gathering and presenting it is not enough anymore.
User needs have evolved
What users of research services now need is
- Getting advice: Being helped at choosing which data will be useful, being explained what methodologies can be used, being shown how the results of the research can be applied to use cases
- Being reassured: In most cases the data will be used to make decisions, it is therefore crucial to clearly understand potential limitations of the research, to have access to credible tools facilitating the decision making and to have a sound understanding of what the data means in its context
- Being challenged: Researchers are no longer required to act as “executants”, on the contrary they add most value when they say “no” to clients when needed. “No” to a methodology or timeline which would not bring results, “no” when another approach would be more cost effective or more efficient, “no” when the research does not confirm what the client was expecting and the client would prefer other conclusions, “no” when the suggested approach is unethical
Research professionals have to re-engineer the research value chain
To be able to cope with these new requirements, research professionals have to completely re-engineer the research value chain. The research cycle should start earlier, get broader and finish later.
In other words from this
It should become this
The research process needs to go through six key steps
Let me explain each of the six steps.
Business objective: The research process should start with an in depth understanding of the research context and objectives. The researcher needs to understand from the client (internal or external)
- What is the strategic context of the research – Are we looking at a growing company willing to invest in new geographies or at an established player challenging its current coverage? In one case we would need to get accurate and detailed results, while in the other a high-level view might suffice
- What has already been done – Has the client tried doing the research themselves him/herself or did they outsource it unsuccessfully? Was a previous version of the same piece done before, and if so, what worked and what did not?
- What is the corporate agenda and the personal agenda of the client – Is there an internal debate around the topic or are we speaking about a subject on which there is a consensus?
Data strategy: The conversation should continue around which data can answer the questions of the client.
- In some cases, the clients know precisely what information they need, in which case the researcher needs either to confirm to the client that the data will answer their questions or to challenge their assumptions and propose alternatives. For example, clients might ask for a customer survey with a sample size of 2000 respondents to enquire about brand image whereas a better approach might be to leverage social listening to provide a much richer view
- In many other cases clients rely on the researcher to tell them what data could answer their questions. In this case it is important to present what data can be obtained, with what cost levels, with what potential limitations and how the data will be presented (an Excel sheet, an executive summary, a short PowerPoint presentation…).
- Using an “issue tree” can be very effective – It would mean starting from an ideal research outcome (for ex. finding an actual figure on the sales of peanuts in the city of Milan), and assessing what “sub-optimal” outcome would also work (for ex. finding the imports of peanuts in Italy and using the share of spend of food in Milan as a way to estimate the peanuts market in Milan)
Research strategy: Then the researcher needs to define the research strategy (some hints can be given to the client during the scoping call or meeting but without a firm commitment). Here the researcher needs to reflect on any methodology which could bring the desired results and finally advise clients on the best approach or approaches, arbitrating between expected quality, cost and timing. A few recommendations for this phase:
- Do it in a team – The researcher should make sure whenever possible to brainstorm with colleagues to generate ideas (no one knows all the possible methodologies and we are more creative when we interact) and to have a colleague review the proposed approach to validate they “buy it” (we are often prisoners of our way of thinking)
- Propose alternatives – When an approach is not obviously better than another, propose both and highlight the pluses and minuses and recommend one; When a research phase depends on a previous one, propose scenarios for phase 2; If the results are uncertain, propose to start with a pilot to assess data availability
- Outline both the time and expense budget involved with the research – Possibly with alternatives (by spending xx hours and spending yy USD we could get AAA, whereas by spending ww hours and zz dollars we would get BBB)
- Verify your approach is MECE (“Mutually Exclusive and Collectively Exhaustive”) – Make sure that every effort done will lead to tangible results, that no effort is redundant with another and that every expected output has a corresponding element in the approach
Research execution: Then, once the client has validated the approach and the budget, the researcher can finally execute the research. At this stage the researcher needs to:
- Make sure to be flexible and consider both primary (field and remote) and secondary research options – Field research gets “new” data, but generally at a high cost, with long timelines and with potential biases; Remote primary research (phone, web) gets interesting results, generally quite fast and at affordable costs, but sometimes with strong biases (for ex. internet access), Secondary research gets pre-processed data, but generally at a lower cost and faster
- Assess all potential resources available to execute the research in the best and most effective way, like leveraging colleagues or free lancers, purchasing reports or mining databases.
- Use “proxies” in case the actual data points researched are hard to come by and proxies would give similar results (for ex. Using the total evolution of a country healthcare spend over 2-3 years as a proxy to the evolution of the medicine spend, assuming the share of medicine in healthcare spend will not change dramatically in such a short period)
- If the scope involves doing several similar pieces (like covering several countries, or several product segments) the research might want to start with a “pilot” to test the approach and to assess the base speed before replicating at a larger scale
- Make sure to share interim results with the client and to get his/her their feedback quickly
Analysis and Packaging. Once the research is finished, the researcher can analyze and package the research results. Here the bottom line is to deliver insights and not only raw data, in other words the raw data means nothing while the insights must answer the strategic questions which were asked initially. This involves the following
- Analyzing the data – After the research, significant value comes from aggregating the data from the different sources, from connecting the various information pieces and from calculations on top of the aggregated data. The researcher should aim for concise and targeted analysis depending on the client questions, for example in order to provide a comprehensive multi-angle assessment of a company, or for describing relevant trends with quantitative data.
- Performing an in-depth quality assurance – This means not only assessing if the data is correct but also if it makes sense, and if doesn’t, then either challenging its correctness or proactively providing explanations reg. these counter intuitive results
- Packaging the results – Depending on how the client wants to use the results, , the researcher should choose the best-suited way of presenting them. If the data is a database of clients with contact details, then an excel sheet is much better than a PowerPoint, on the contrary if the results are to be presented at a board meeting, then slides are on point. And the sources always need to be added in the documents !
- Deriving insights – The researcher needs to explain the meaning of the data. If the client wants to understand a new phenomenon an executive summary with the key learnings is a must. If the client needs to choose between countries or partners to do business in/with then a simple excel model allowing to score them is great.
Handover. The research doesn’t have more value than what the client understands and remembers of it this is why making sure the client gets a proper handover is crucial. Here are a few good practices.
- Structuring the answers as per the clients initial questions – So that the client can easily absorb the output of the research
- Handing over to the client – The client owns the research, he/she they must be able to get full transparency with regards to the sources of the research and how the research was conducted. The researcher therefore needs to transfer all the documentation used during the project (interview notes, backup files etc.), and provide explanation about all the content which has been generated. In particular it is important to explain the potential limitations of the data and the potential biases of the approach.
- And of course, asking for feedback once the work has been delivered
NB: It is not because the number of steps is doubling that the time to execute will double, in fact it will probably go down, allowing the researcher to make the proper trade offs in the best interest of the requestor.
We need to move to a consultative approach to research
If this approach is followed, which I would phrase as “consultative approach to research”, I can guarantee that the research profession will vastly increase its level of value add, which will in turn reinforce the interest of the job and generate many new opportunities !
Your comments are welcome !
If you want to hear more about how Infomineo supports its clients using this methodology you can reach out
As written, research roles are evolving into higher value added roles but there is the perennial struggle with organisational leadership to see research beyond a support function, thus limiting growth opportunities for great performers. The will to evolve should be the common goal.
Public Policy & Strategy Consulting
5 年Great piece ????
UX Research | Project Management | Data Analyst | Qualitative & Quantitative Research | E-commerce | Insights & Strategy
5 年Interesting read Martin!! Thanks for sharing the knowledge.