In search for a data-informed way to build product vision and positioning
Focusing on a customer decision to build positioning and product vision
One of the most memorable working experiences came from significant research undertaken with my colleagues at one of the companies I worked for and a consulting firm MarketFit, led by Alan Albert. It was about product vision and positioning — a not-so-easy topic for even the most experienced product managers.
Many product managers and product marketing managers do not realize how they should tackle crafting product vision and positioning in their company. The problem seems to be covered with mystique.
At best, companies run internal workshops and brainstorms to devise product vision and subsequent positioning based on their understanding of the market, customers, and product they are building.
Business books, including Blue Ocean Strategy, describe the same way to get at positioning or product vision. The implication is that people working on a product should know enough collectively to devise a product positioning. While this approach is better than a top-level manager making a sole decision, we could do much more to get product vision and positioning that reflects what our market segment truly needs.
We typically position what we create instead of creating what is easy to position
In general, positioning follows a product vision. After all, we start with a vision and only then get to the level of details: what features do we build, what positioning our product will have, how will we appeal to the needs of the user, how do we make sure our potential audience realizes that we fit their needs.
However, we should not start working on positioning after we deliver a product or a version of it. If we start with both product vision and positioning, we will be able to focus our efforts on features that the target audience will appreciate. Some features will not make the cut, and that is fine.
The typical way to do product vision and positioning isn’t optimal
Product vision, positioning, and value proposition often come as a result of a sole or group decision. Managers write positioning statements and documents and then push those documents to GTM teams.
A sole decision-maker introduces too many biases
Industry, market, and client knowledge is nothing to sneeze at, but our experiences within an industry or even insights from tangential interviews introduce biases to our thinking, and they shape the way we see our product and the needs of our existing or potential customers.
A group comes to a compromise between multiple biases
Collaborative work, when multiple people exchange opinions and provide ideas on product vision and positioning, is a much better alternative to making a decision based solely on the gut feeling of one person. Triaging multiple data points results in a much more consistent and relevant output, although the biases are still present. You just remove the most egregious outliers by doing the work together. The result is a compromise between the biases of different people.
Sometimes the typical way works — we nail the positioning, product vision, and value proposition. The problem is not that we can’t get to an effective state with this approach. The problem is that the typical way is not consistent and depends significantly on luck, even when we include experts. So many products fail despite the names behind them.
A data-informed way to build positioning
Within this series we will go through a series of steps towards getting a data-informed product vision and positioning.
What’s so different from the ‘typical way’ described above?
First of all, instead of guessing what points we should highlight for our target audience, we’ll take those points directly from those who have recently made a decision on selecting a solution like the one we are building.
Secondly, we will have to say ‘no’ to things that don’t align with the decision-drivers of our audience. Cool ML algorithm that you can build and introduce that will make you stand out (as it covers some specific need), but that’s not relevant for the audience you are building your application for? That’s a no. However, you might want to think about finding an audience that cares about that but the burden of validation is still present.
Thirdly, we will not just do the interviews to get the decision-drivers, we’ll also quantify them — through interviews.
As the main idea of the series is to show you how to get a product vision and positioning based on the decision-drivers, that’s where we’ll spend the most time on. For target audience definition, ICPs, roles within B2B I suggest you read an additional article or two if that’s not defined or used within your organisation.
Step 1: What decision does your target audience need to make?
Define an ICP first
At the outset, you need to have an idea of what the most fitting segment looks like for your new product introduction. You may be wrong at that point, and potentially the interviews will show you that you can’t cover decision drivers of the audience that you have thought of as the best for you, but a rough image of an ideal customer profile is necessary.
A problem you are solving with your product is felt the most by a specific category of people. Who are they? What do they do day-to-day? What kind of job role within the company do they have, if that’s relevant? Create a portrait and try to screen participants through it in the next steps. For B2Bs you’ll start with the company description first, and person description second — you will need both.
If you have no experience in defining a persona or ICP, I suggest you find an article about that or a person within the organisation that can help you. Don’t skip that part just because you think you know your audience. A preliminary research on segments of your target audience may be necessary, or you may use existing data to build your ICP.
Segment definition will impact all the work you are going to do next! Different segments will have different decision-drivers.
Restate the decision into a question
The more important point is to know what kind of decision your audience needs to make. A decision can generally be restated into a question, a simple example for B2B SaaS would be: “What product from [product category] do we want to select?”
Specific needs, values, criteria, and motives drive an answer to this question. The next steps will shed light on the specific decision drivers that help your target audience answer that question.
Restating the decision into a question is my preference, not a hard rule. Knowing the decision is enough, I just feel like it’s much easier to empathise with the audience a bit more when the decision is restated into a question.
Define the role to talk
There are multiple roles within B2B SaaS solutions: decision-makers, evaluators, end-users, and others. Sometimes they overlap, sometimes they do not — for our purposes, we need to select which roles drive finding the answer to the above-mentioned question.
For one research, you might realize that evaluator and decision-maker will be different people within the company, but evaluator will be much more crucial than a decision-maker, just because decision-maker approves or rejects what evaluator suggests. In other cases, an evaluator will provide a selection of options a la carte for a decision-maker to select from, and in this case, you need to decide: do we want to get on the list or do we want to get selected from the list that we will be a part of.
An imaginary example for the series
For example, you are creating a B2B SaaS application that allows your clients to accept payments on a website / in an application.
The decision of the target audience, in this case, is related to a question: “Which payment provider should we introduce to our product?” I am omitting a lot of details, such as the specific segment of the target audience, ideal customer profile, vertical or exact use case for getting payments, etc, that I described previously. In your research, you’ll need to have much more clarity on those points.
This concludes the first part. In the next chapter we’ll go about actually gathering decision-drivers and their rankings as a foundation of the positioning and product vision.
By that point we know:
– What our Ideal Customer Profile (ICP) is for the product we are considering (or for a problem we are trying to solve)
– What is the decision to be made by an ICP or a representative of ICP (or what question they are answering)
– What role should we talk to (for B2B)
Step 2: Discovering the decision drivers
In product management, we run multiple interviews about problems, solutions, and context, we use different methodologies or hire research agencies to do the footwork. But while we do that for so many different reasons, interviewing to get specific decision drivers is not on top of the list for many product managers.
From here on, I’ll describe the high-level overview of the research process for a product that hasn’t seen the light of the market yet (although you can do it for an existing product too) but is going to be in an existing product category.
This step is about digging wide, not deep. The first thing you need to do (omitting the customer qualification process for the interview, screeners, etc) is to understand the drivers of a decision you care about.
When selecting something, both for business or personal usage, we are guided by specific motivations, needs, and criteria. Sometimes those criteria become explicit (when we create a list of pros and cons), sometimes those criteria stay implicit (when we decide to buy a new shiny thing, there still are specific attributes, even such as ‘my friends have it, so I should too’, that drive the decision-making process).
The main idea is simple — by asking questions that relate to a past decision (‘What if we are creating a completely new market category?’ will be discussed in part 3), find out and make explicit all of the decision drivers.
At this step, you are just looking to get all those drivers, context on them, exact meaning behind each driver, the overall process for the made decision, who was involved, what sources of information were used, who was consulted from interviews.
After conducting interviews, you will need to consolidate the list if you get too many decision drivers. It could be a bit excruciating because you take your interviewee’s words and create a new decision-driver by grouping multiple things into one. If you have the ability, validate with the interviewee that what they said could be accurately represented in a group that you have created in the way you named it.
Example questions for this stage:
In your process of selecting a [product category], what was the first thing you did in your research phase?
Who were stakeholders in this selection process and what was their input / what they cared the most for?
What made you select your current [product category] provider?
How did you compare products from [product category]?
Which products from [product category] have you considered in your research process?
As a follow-up: How does the selected [product category] provider differ from all other providers?
As a follow-up: What was the second-best solution as a [product category] provider? What made them a bit worse than your selected [product category] provider?
The outcome of the interviews that you’ve run is a distilled list of decision-drivers. It’s quite probable that decision-drivers will differ across the subjects, and even the same decision-drivers will be called differently by different people.
You will need to analyze results and join decision-drivers that are similar. This is a subjective step — you will have to go through interviews to understand whether certain drivers were really similar based on the context provided around them.
Next, shed the list. You want to have no more than 10 or 15 decision-drivers for the next phase.
By that point we know:
– What our Ideal Customer Profile (ICP) is for the product we are considering (or for a problem we are trying to solve)
– What is the decision to be made by a representative of ICP (or what question they are answering)
– What role should we talk to
– List of decision drivers
Step 3: Ranking the values that drive the decision & getting the context
With the list of drivers that were used for selecting a product in your category you can go to phase 2 — getting the top drivers, so you can build product vision and positioning for your product.
There are a few ways to go about ranking decision drivers. The general idea is the same in both scenarios — you want to get a quantified rating of each decision driver.
A question you can use is “Please, rate each expression on a scale from 1 to 7, where 1 means ‘completely disagree’ and 7 means ‘completely agree’.”, where each decision driver becomes part of a sentence “In my process of selecting a product from [product category] [decision-driver(i)] was a key decision factor.”
Ask for ratings within the phase 2 interviews…
The first way is to ask for ratings within each interview. This will require asking the same question about different decision drivers multiple times, while an interviewee has to sit and think about whether he/she agrees with the statement or not.
Some interviewees may feel pressured, some will think for a long time — it’s a time-consuming process. The biggest benefit of this scenario is that you can encourage the interviewee to use the whole scale if you see that he/she started to answer just 1-s and 7-s.
… or do a survey before the phase 2 interviews
If your typical interviews run for too long, you may want to consider running a survey before doing a second phase interview.
The question stays the same but is now presented to an interviewee 1 or 2 days before the interview. He/she will be able to think through it without any pressure, and you will not spend any time assigning ratings during the interview, which allows you to ask more questions about the context of the most important decision drivers.
Our goal at this stage is to understand the five most important factors for each interviewee, then get a context on each of those.
The context is the king — five dimensions of a decision-driver
The context will enable you to create product vision and positioning that speaks directly to your target audience. The ratings will help us understand what we should learn, while the context will allow you to set your understanding straight on what the decision-driver means, how it is typically uncovered, how it is measured. All those things are inputs to the phase where you’ll actually be crafting product vision and positioning.
I propose five dimensions that you should learn about the decision-driver no matter what else you want to learn. Sometimes you’ll feel that the dimension is not applicable — you should still ask the question, just in case you learn something insightful or surprising. Let the interviewee tell you that the question doesn’t make sense instead of censoring yourself from asking the question.
Example questions for each of the dimensions:
Meaning: How do you understand [decision driver A]? And what wouldn’t be possible if [selected solution] didn’t have [decision driver A]?
Indicators of presence: How did you know that [decision driver A] was present in your [selected solution]?
Measurement: How did you measure how much [decision driver A] is covered / present with [selected solution]? Or How did you learn that [selected solution] will have an acceptable level of [decision driver A]?
Comparison: How did you compare your selected solution with other solutions in the category in terms of [decision driver A]?
Indicators of credibility: What made you confident that [selected solution] will cover [decision driver A] going forward?
When asking questions about those five dimensions, keep the question the same for all decision-drivers because that will allow you to gather context in a structured way without introducing too much or too little information about specific decision-drivers, thus biasing the research.
The more context you get, the easier it will be to build a positioning because you will not only know the decision-driver but also the meaning behind it. Thus, you will be able to craft messaging, collaterals, UI texts, emails, in-apps with this context in mind.
By that point we know or have:
– What our Ideal Customer Profile (ICP) is for the product we are considering (or for a problem we are trying to solve)
– What is the decision to be made by a representative of ICP (or what question they are answering)
– What role should we talk to
– List of decision drivers
– Set of transcripted interviews
Step 4: Calculate averages and medians for decision-drivers to get an overall top 5
Hopefully, the interviews that you will run will provide more insights than you had before running the research. Understanding decision drivers is crucial, but we still have a bit of work to do before going back to the ivory tower to craft product vision and positioning.
At this step, we need to calculate average, median, and standard deviation scores for each decision-driver.
Average and median will give you a prioritized list of drivers to use in your product vision and positioning. You will need to see how those drivers fit what you are developing, where you can implement features to support the needs of your audience, and how to better appeal to the audience to signal that you cover their needs.
Standard deviation is an important component that provides a glimpse into how homogenous the responses are. High standard deviation potentially means that you might have different segments within your group of interview subjects because high deviation signals that some respondents think it is an important driver, while others don’t consider it important at all.
Keep an eye on decision drivers with a medium or low average and/or median, but high standard deviation. Potentially, there is a sub-segment to explore further. And for that sub-segment, you could potentially create additional landing pages with messaging exploiting their decision drivers, if you understand that sub-segment well and can target it with specific advertising.
If you have hypotheses beforehand, for example, your product can be both used by B2C and B2B companies, but you believe their decision process is different, you can calculate the above-mentioned metrics separately for those two audiences as well to see how your hypothesis holds up. If there are indeed strong differences, you might want to think about catering to just one audience (either B2C or B2B in this example) in the beginning. And in B2B there potentially is a huge difference in decision drivers for companies of different sizes.
In addition to calculating the metrics and ranking the overall results, write a description of each value along with the context that you were gathering. If you’ve asked what the driver means for the customer, provide a brief description from a collection of answers that is supported by direct quotes that you’ve gathered. Do the same for each question that you’ve asked about a decision driver.
By that point we know:
– What our Ideal Customer Profile (ICP) is for the product we are considering (or for a problem we are trying to solve)
– What is the decision to be made by a representative of ICP (or what question they are answering)
– What role should we talk to
– List of decision drivers
– Set of transcripted interviews
– Ranked decision-drivers with supporting context across all interviews
Making product vision and positioning from the research
At this step, you are going to start crafting product vision and positioning from the information that you’ve gathered.
Features enable you to describe the values of your product that reflect the decision drivers of your primary customer segment.
Select no more than 3 decision-drivers at this stage. The more you try to address, the fuzzier your product will be at an early stage.
It is probably obvious from the previous steps that your product vision should address top decision-drivers, and should reflect how your product will make it easy for the target audience to select your product among others.
Feature-wise, use decision-drivers as a prioritization criteria for the features you are considering to build. If a feature doesn’t address important decision-drivers, you must have a very good reason to build it (technical necessity and basic expectations are good enough reasons).
It is no less important to consider how you implement the features as many solutions may have the same feature, but the way it is implemented in one solution will address the decision-driver, while the way it is implemented in another solution will miss the mark for the target audience.
With product vision and decision drivers at hand, you can build how you are going to position your product alongside competitors. Appealing to all decision drivers upfront may not be the best strategy, and the better way is to double down on 1–3 decision drivers, while 1 or 2 more can be covered in your positioning as supporting points.
A word of caution: your positioning is how you are being perceived. Whatever you are going to say about yourself in the market, whatever you are going to write down in your internal strategy memos, if it’s not reflected in how customers understand your product and perceive the value of it, it’s just not going to be your positioning. Customers will listen to your marketing messages but in the end they will be the judge and jury of your position, so do your best to stay consistent both in your marketing and in your product.
The ‘how’ on actually writing positioning or describing the product vision is a burden I will have to leave with you, as this is not the main topic of the article.
Scaling it up
As (if) the business grows, you are going to be looking to expand your target audience. The methodology still holds, and you can approach scaling up from two different perspectives:
Reconsider the decision drivers that you’ve gathered and start covering more of them either within the main product or as an add-on
Perform the research once again, but with a specific segment in mind
In the first case, the idea is to use the data that you have already gathered to see what decision drivers can be covered that potentially unlock a bigger chunk of the market. An easy example here is advanced compliance with multiple regulations (if that was a decision driver for your audience, but wasn’t high enough on the list) if you are in a product category with different levels of data sensitivity. A company that has it as a decision driver just can’t select you, because it’s simply a blocker. You need to study the companies that rated it high, just to see what else they require. Is it enough for you to cover the decision drivers that you initially covered and do you just need to cover one more? In that case, it seems like a good opportunity.
If you conducted the research a long time ago, the results may not be applicable, so approach it with care.
In the second case, you just go through the full cycle of research once again and just follow the outlined steps.
What if you are creating a completely new market category?
The whole article was dedicated to building product vision and positioning for a product in an existing market category. A fair question is whether we can follow this methodology to create the same artifacts for a product that seems so novel that it is going to create a completely new market category?
In short, as related to this specific methodology, I don’t know — I haven’t tested it on completely innovative products. Regardless, I’m going to speculate about that.
In theory (I warned you I’m going to speculate here!) you could start with a problem for a specific target audience, rather than a product or product category. But in this case, what exactly should we research for building our product vision and positioning?
Each existing problem has a solution. Sometimes the solution is to do nothing, and sometimes there are workarounds employed by people or companies who feel the pain of the problem. We can study how the problem is currently solved, and do it in two ways:
Find decision drivers for their current solution (which you believe to be not optimal)
Find gaps: what exactly people or companies you research are unhappy with? What prevents them from solving the problem completely?
Essentially, the research will be split into two parts, because there are now two lists to be ranked. Follow the same steps, but in the end, you’ll be building product vision and positioning not only based on what they value in their sub-optimal solution, but also on what they miss in the current solution.
Can it work for an existing product?
The challenges of working with an existing product don’t lie in the research area.
Instead, your challenge will be multiple stakeholders that need to be involved as well as sheer organizational effort to use the insights to build product vision and positioning, as well as use them later on in product management, marketing, and sales.
You might run into executives that will dismiss your research just because ‘some research was done already’ or something along the lines. It’s just not going to be a priority for many executives. It will be your work to prove that the research is necessary.
Going back to the research dimension, working on an existing product will provide additional insights because you will be able to talk to your users or decision-makers from existing accounts. Those interviews should be run separately from the interviews with users/decision-makers of competitive tools, or at least separated within the analysis stage. But you will not only learn what the specific segment of the market cares about but also what were the decision drivers for the existing accounts as well.
Gratitudes
Thanks to Alan Albert, MarketFit president, who has decades-long expertise in building products, crafting positioning, and optimizing pricing and led me to dive deeper into discovering customer values.
Thanks to Dmitriy Amosov, Sinisa Kravarascan, Maja Blazek, Tomislav Jus who were an integral part of the initial research that was a foundation for my learning of decision-drivers.