One option to drive this product discovery work is to make use of a psychological mannequin known as an Alternative Answer Tree (OST). To assist product groups successfully uncover and seize worth, we’ll stroll by what a possibility answer tree is, easy methods to use one, an instance of an OST, and easy methods to join your OST to analytics and experiments.
Key Takeaways
- Product discovery combines qualitative perception and quantitative analytics to find out high-value buyer must be solved.
- Alternative answer timber assist product groups concentrate on key areas of buyer want:
- Metrics constrain discovery to business-relevant areas
- Alternatives determine buyer pains to unravel
- Answer concepts present hypotheses for creating buyer worth, and
- Assessments and experiments validate or disprove answer concepts.
- By connecting the visible framework of alternative answer timber to an analytics and experimentation stack, product groups can ship worth extra rapidly.
What’s product discovery?
Product discovery is the set of processes that groups use to unveil and make clear the wants of their prospects. By understanding the sorts of ache factors that prospects expertise, we will prioritize the sorts of options that may create worth. Product discovery is an important subset of the general product improvement course of, because it helps product groups make knowledgeable choices about which product concepts to pursue and which to put aside.
Product discovery consists of buyer interviews, buyer shadowing, surveys, buyer suggestions, and prototype testing. Product discovery ought to occur commonly and regularly. In different phrases, discovery shouldn’t simply occur on a quarterly foundation; quite, it ought to happen on daily basis, inside each product pod, in small ongoing increments.
Buyer-focused product discovery allows us to maximise our understanding of what creates worth for purchasers, and to replace our strategy in actual time as we constantly study.
However, whereas product discovery is essential to changing buyer wants into enterprise worth, this important work is typically criticized or deprioritized by executives and cross-functional stakeholders. Let’s dive into why this may occur.
Why may product discovery run into adoption challenges?
Whereas most leaders and stakeholders received’t object in precept to product discovery, many product groups nonetheless commonly run into adoption challenges with their product discovery practices.
It’s because many leaders are uncomfortable with undirected discovery; that’s, too regularly, product groups will embark on a “discovery safari” to “take within the sights,” which causes leaders to query whether or not this analysis is sufficiently focused to drive enterprise worth.
Particularly, leaders need to have solutions to the next sorts of questions:
- How do we all know that discovery will transfer a enterprise metric that issues?
- How lengthy is that this discovery work going to take, and the way a lot does it price?
- How do we all know that this funding will repay?
In our teaching observe at Product Trainer, we discover that many product managers initially fail to deal with these questions upfront, inflicting pointless friction and pressure. We’ve discovered that by introducing the chance answer tree to our shoppers’ working processes inside their organizations, product groups are considerably extra seemingly to reach advocating for customer-driven discovery.
So, let’s dig into what alternative answer timber are, and the way they ship worth throughout all elements of the group.
What is a chance answer tree (OST)?
The chance answer tree framework was initially designed by a Stanford College design professor. In 2016, Teresa Torres, founding father of Product Speak, utilized this framework to product discovery processes.
Each alternative answer tree accommodates 4 key elements:
- Metric: the business-relevant metric that guides discovery
- Alternative: the ache factors that prospects have
- Answer concepts: the potential ways in which we as a product group can deal with buyer ache
- Assessments: the experiments we will run to validate or invalidate our options, enabling us to de-risk our options and swiftly ship iterative worth
Under is an instance of what a possibility answer tree may seem like. We’ve coloration coded metrics blue, alternatives inexperienced, answer concepts yellow, and experiments orange. Don’t fear about studying the textual content for every. We’ll dive deeper into this instance in a later part.
Why did Torres create the chance answer tree mannequin? She did so primarily based on her observations in coaching product groups by product discovery. She observed that groups wanted a visible construction for guiding their discovery efforts, aligning proposed options vs. found buyer pains, and securing buy-in throughout departments.
Torres drew on a way taught by Stanford professor Bernie Roth. Professor Roth requested how folks’s desired options linked to their underlying wants and ache factors, after which requested individuals to conduct “divergent solutioning” to provide you with completely different potential options for fixing the identical ache factors.
By framing this tree-like set of questions as a visible graph of a possibility answer tree, Torres discovered that product groups have been more likely to advocate for options that really addressed the ache of their prospects.
How does a possibility answer tree enhance discovery adoption?
Crucially, the chance answer tree visualization does the next for product managers, government leaders, and cross-functional stakeholders:
- Establishes a business-critical metric as a focusing lens for product discovery
- Focuses efforts on fixing buyer ache quite than constructing a guidelines of options
- Converts discovery insights into “alternative areas” to spend money on
- Shifts the dialogue away from “function supply” in direction of “fast experimentation”
- Ties experiments again to buyer ache and iteration alongside prospects, quite than non-interactive inner brainstorming
Earlier, we talked about that leaders are eager on understanding the enterprise worth that product discovery may deliver, and the potential prices or dangers related to product discovery. By framing discovery into business-relevant outcomes, product groups can ease fears round whether or not product discovery insights can be actionable or not.
And, with a possibility answer tree in place, groups are a lot much less more likely to over index on a given answer. As a substitute, they’ll take the time to ask what alternative or buyer ache they’re fixing, and use this broader view to innovate and provide you with higher-impact, lower-cost options that in the end create extra ROI.
The visible construction of the chance answer tree clearly reduces the precedence of any given function because the “finish consequence,” and as a substitute strengthens the message that shifting the business-relevant metric is what actually issues.
Subsequently, product groups that efficiently introduce alternative answer timber to their orgs have a tendency to search out that stakeholders are way more keen to run with experimentation, quite than specializing in a laundry checklist of options with deadlines.
How do alternative answer timber work?
Let’s break down every of the 4 key elements of alternative answer timber, and focus on greatest practices for every.
Metrics
The metric in your alternative answer tree must be aligned with a KPI (key efficiency indicator) out of your OKRs (targets and key outcomes) in order that your work is immediately tied to enterprise success. This metric is the lens by which you’ll conduct your product discovery.
In different phrases, any buyer conversations or product discovery efforts that may not transfer the metric shouldn’t be thought of. Solely the initiatives and efforts which have an actual chance of fixing the metric must be actively investigated.
However how can we choose a superb metric? Good metrics ought to steadiness the stress between “enterprise success” and “proximity to product.”
If the metric is just too distant from the product (e.g., companywide income or companywide income), then the product group will wrestle to drive any form of actual focus of their discovery efforts. Theoretically, any initiative may very well be justified as a option to improve income or lower prices.
And, if the metric is just too near the product (e.g., function click on charges), then the group is over listed on a selected answer, and never targeted sufficient on buyer ache or funding alternatives that may transfer the needle for the enterprise.
metric ought to observe that worth has been created for purchasers, and that worth has been captured for the enterprise. Examples may embody time on platform, or month-to-month lively customers.
Alternatives
Simply because we all know what metric we’re going after, doesn’t imply that we all know what pains our prospects are experiencing. Many occasions, waterfall-oriented groups provide you with concepts “contained in the 4 partitions of the corporate” with out studying from prospects about what their pains are.
Alternatives must create new worth for purchasers, whereas additionally enabling the corporate to seize that worth within the type of utilization, income, referrals, and different related enterprise drivers.
Product groups can leverage any of the under to determine alternatives:
- Inbound buyer suggestions
- Outbound 1:1 interviews
- Outbound surveys
- Product knowledge analytics
- Discussions with inner stakeholders (e.g., assist, advertising and marketing, gross sales)
To make this idea extra concrete, let’s contemplate Gmail, which is an e-mail inbox supplied by Google.
A chance is not “I would like a greater spam filter.” It is a answer thought, and we’ve fallen for the entice of listening for what to construct quite than which pains to unravel.
Once we contemplate the function thought “I would like a greater spam filter,” this may really level to a wide range of completely different underlying person pains, resembling:
- It takes me too lengthy to discover a particular e-mail
- I don’t like receiving emails from folks I don’t know
- I’m pissed off that e-mail notifications hold interrupting me throughout shows
As you may think, the set of potential options to deal with the chance of “discovering an e-mail sooner” vs. the set of potential options for the chance of “stop emails from interrupting me” are wildly completely different from each other. Whereas it’s true that they share the identical function thought of “I would like a greater spam filter,” that doesn’t imply {that a} spam filter is the easiest way to attain both goal.
That’s why it’s essential for us to conduct qualitative buyer interviews and quantitative analytics to grasp the true ache that’s being expressed behind any given function request.
Moreover, any time our colleagues suggest options, we must always not take the function at face worth. As a substitute, we must always ask “what underlying ache does this function request clear up for a buyer?” From there, we will then contemplate the complete set of answer concepts, lots of which is likely to be extra satisfying, simpler to construct, and simpler to keep up than the preliminary proposed thought.
Answer concepts
When you’ve chosen a possibility, accomplice with design and engineering to determine a wide range of potential methods to assault the client drawback:
- Be divergent and generate as many concepts as you’ll be able to
- Don’t be afraid of “dangerous” concepts as a result of they’ll spark good concepts
- You’ll converge on “what to ship” later throughout experimentation
Design can assist you contemplate how folks at the moment deal with this drawback in the true world. They will assess how rivals clear up this drawback by their merchandise, and so they can even determine how prospects use substitutes, alternate options, or guide processes to alleviate their ache. This 360-degree view allows us to provide you with significantly better concepts for options.
Engineering can assist you establish whether or not this buyer ache will be abstracted right into a broader ache that encompasses a number of associated use circumstances. In some cases, many pains may appear fully unrelated at a person degree, however is likely to be simply solved in a single fell swoop at a system degree.
For instance, say that you simply’re in command of a mission administration platform like Asana or Trello. Some customers need to know which duties are prone to lacking their deadlines. Some customers need to know whether or not some assignees are overburdened with work. Some customers need to perceive which departments made probably the most job requests. And a few customers need to know which teams have completed probably the most duties.
These all appear wildly separate from a person ache perspective, however engineering can determine that the underlying person ache is “we will’t simply group or filter issues collectively primarily based on attributes that we have already got about every job.” And, from that lens, engineering can suggest a extremely versatile sort-and-filter system that may not have been instructed by any single person.
Lastly, don’t be afraid to solicit suggestions from cross-functional companions! In any case, they might have function requests available so that you can contemplate.
However, as we talked about earlier than: when participating with cross-functional suggestions on answer concepts, you’ll want to make sure that these really match the chosen alternative space that you simply’ve prioritized. You don’t need to merely deal with the answer concepts that may assist shut a deal or fulfill a very vocal buyer.You need to ship probably the most potential worth, utilizing the fewest assets obligatory, within the quickest potential time.
Assessments & experiments
To attain excellent ROI, answer concepts shouldn’t be absolutely shipped as-is, as a result of that’s costly and high-risk. Inside every answer, now we have a number of underlying assumptions–and, we must always take a look at every of those assumptions by well-designed low-effort experiments. This strategy reduces danger, drives studying, and in the end will increase ROI.
For instance, say that you simply’re in command of a contract software program platform like Ironclad, and also you’d wish to ship performance that robotically plugs in related contract phrases primarily based on AI/ML. Relatively than constructing your entire advice system in a single swing, you’ll need to de-risk this initiative by breaking it down into smaller items.
One assumption is likely to be “we imagine that legal professionals need to evaluation the suggestion earlier than inserting it into the contract.” If that’s the case, then the UI that we ship ought to let legal professionals see the contract clause that’s being instructed, and provides them the flexibility to simply accept or reject.
But when our assumption is “legal professionals are comfy with software program filling this in for them, and having to have a look at dozens of apparent recommendations is a waste of their restricted time”, then we shouldn’t construct this UI.
How would we take a look at this? We may run a “particular person within the machine” take a look at with a lawyer to see whether or not they’re comfy with contract clauses being inserted with out them taking a look at it.
One other assumption we would want to check is “contract clauses are constant throughout completely different sorts of contracts.” This won’t be true. Totally different contracts could require clauses to be edited earlier than inserting. In that case, we’d additionally need to run an experiment right here.
For every experiment, we have to guarantee that we’ve carried out the next:
- Establish all of our implicit and specific assumptions about our prospects, our rivals, and the expertise that we’re utilizing
- Create upfront thresholds for “how a lot the metric ought to transfer” for us to say that it was profitable
- Draft predetermined subsequent steps primarily based on experiment efficiency, protecting what you’ll do in case your assumptions are validated, in addition to what you’ll do in case your assumptions are disproved
How can we prioritize the completely different experiments that we need to run towards our answer concepts? Usually, we need to prioritize with these three elements in thoughts:
- Which experiments will transfer the overarching OST metric probably the most?
- Which experiments will de-risk the most important assumptions?
- Which experiments will price us the least effort and time?
When to make use of a possibility answer tree
The purpose of a possibility answer tree is to arrange the potential areas of alternative the place the corporate can make investments; in different phrases, it categorizes viable ache factors that we will clear up for our prospects.
Subsequently, alternative answer timber must be used as an enter into any form of roadmapping train, whether or not we’re constructing roadmaps at a quarterly degree or an annual degree. In any case, we will’t construct efficient roadmaps if we don’t know which ache factors our prospects have.
Alternative answer timber are additionally fairly helpful for framing product proposals, particularly after we plan on presenting these proposals to government groups. By making certain that we’re aligned with our executives on the important thing metrics that matter, and by offering an area to collaboratively uncover the completely different alternatives obtainable to us, we will extra successfully iterate by potential options and ship extra worth with much less effort and fewer time.
Many product groups run common product technique offsites. The following time you attend a technique offsite, contemplate pitching “alternative answer timber” as an actionable subsequent step for all attendees after the offsite has concluded.
And, for future technique offsites, contemplate asking organizers to start with a evaluation of any present OSTs that groups have been actively engaged on. By doing so, all attendees may have a shared view of the chance house forward, in addition to the progress that’s been made on tackling these alternatives.
An actual-world instance of a possibility answer tree
To deliver the idea to life, let’s contemplate a real-world instance of alternative answer timber in observe. Let’s say that you simply’re liable for enhancing the search functionality inside your B2B product, which is a web site builder (e.g., Squarespace, WordPress, Wix, and so forth.).
You might need outlined your metric of success to be “the common variety of searches accomplished, per week, per buyer.” That’s, in case your search functionality is offering worth to your prospects, you then would count on that your prospects’ finish customers would more and more use your search performance, quite than the performance of another competitor answer (e.g., Google Customized Search).
By means of person analysis and inbound buyer suggestions, we would have found that there are three core alternatives for enchancment:
- Search outcomes inside your function are at the moment fairly troublesome to sift by
- Customers have to attend a very long time for search outcomes to come back again, which is irritating
- Many customers are including “native” location-based key phrases (e.g. “New York” or “San Diego”) to their queries, which is repetitive and tedious but obligatory for correct outcomes
We’d provide you with a possibility answer tree just like the under:
You’ll have labored by discussions with gross sales, advertising and marketing, buyer success, buyer assist, design, and engineering to determine potential answer concepts for every of the three alternatives.
And, for any given thought, you don’t merely ship that concept from begin to end – you create testable hypotheses and experiments to determine which facets of the concept are most useful.
For instance, maybe you need to make it simpler for customers to sift by search outcomes, and one concept that appears compelling is to indicate consequence previews to customers.
“End result previews” is kind of broad, nevertheless; what precisely makes a preview compelling?
Say that we need to take a look at the belief that “each consequence must be reviewed by customers.” In that case, we might spin up an experiment the place each returned consequence has its personal text-based preview related to it.
However, we may additionally take a look at the flip facet of this assumption, the place “a person actually solely simply desires a single greatest consequence ASAP.” We’d additionally need to run an experiment with a single highlighted “most related” textual content snippet on the high of the web page.
And, how do we all know that textual content is actually the easiest way ahead? What if our customers are significantly keen on pictures, much like Pinterest or Instagram? In that case, as a substitute of surfacing text-based hyperlinks, we must always run an experiment the place we floor image-based snapshots for our search outcomes as a substitute.
The interplay between OSTs and product technique
Alternative answer timber work together with product technique in two key methods.
First, the OST methodology enhances and enhances product technique. In any case, each OST begins from the product technique, as the chosen metric for any OST should align with the recognized product technique.
Second, alternative answer timber additionally inform and affect product technique, primarily based on assumptions that we’ve validated or disproven, and primarily based on the newly-discovered insights that we’ve uncovered from our experiments.
As a rule of thumb, refine your product technique after each three experiments. Ask these questions regularly as you full every experiment:
- Is the recognized buyer phase nonetheless the appropriate one to pursue?
- Is the chance bigger or smaller than we initially believed?
- Has the market change into simpler or tougher to strategy?
Aligning and collaborating with stakeholders by OST
Individuals make choices primarily based on the context that they’ve, and cheap folks make the identical choices, if they’ve the identical context.
If you happen to’re confronted with disagreement, one (or each) of those is going on:
- The disagreeing particular person doesn’t have your context
- You don’t have the context that the disagreeing particular person does
We will use alternative answer timber to deliver stakeholders and executives with us on the invention journey, ranging from buyer learnings and ending with our conclusion on what to ship subsequent.
Let’s break down easy methods to arrange alternative answer timber inside your group. This playbook relies on the group workshops that we’ve run with Fortune 500 corporations, in addition to the 1:1 government teaching that we offer to high-potential product leaders.
Under, we’ll cowl:
- The way to set up your first OST
- The way to mature an present OST
- The way to pivot in direction of a brand new OST
Establishing your first OST
Any time we introduce new working processes or frameworks for considering inside our organizations, we have to safe the buy-in of leaders and cross-functional counterparts.
To interrupt down this support-building step, we’ve discovered that probably the most profitable product managers have a tendency to make use of not less than 4 conferences to align stakeholders in direction of product discovery practices. Listed below are the 4 conferences, and the matters to cowl inside every:
- Metric alignment: Align on the important thing enterprise metric and focus on how discovery will result in higher outcomes than waterfall over the long term.
- Alternative exploration: Stroll by an preliminary set of alternative areas primarily based on buyer suggestions, and invite stakeholders to share their alternatives as effectively.
- Alternative choice: Collectively choose a possibility to start out with (not which one is “proper” or “flawed”), and doc your decision-making course of collectively.
- Answer ideation: Collectively ideate a wide range of options and agree which preliminary answer to discover by experiments.
You don’t need to rush by these conferences, as a result of the journey is extra necessary than the result. If you happen to discover {that a} explicit stakeholder is silent, adversarial, or just confused, take a while to interact with them one-on-one outdoors of those conferences. Be taught what their considerations are, and work in partnership with them to deal with these considerations.
And, it’s possible you’ll want greater than 4 conferences to deliver stakeholders with you on the journey. That stated, we don’t advocate cramming a number of matters into the identical assembly, as that tends to trigger fatigue, loss-of-focus, and frustration from stakeholders.
As an apart, If you happen to’d wish to study extra about easy methods to navigate stakeholder preferences extra broadly, we’ve recorded a one-hour lecture as a part of our month-to-month recorded PM lessons.
Again to the subject at hand: we now have our preliminary alternative answer tree in place. However, we’re not carried out but.We nonetheless must provide you with experiment concepts, and we have to share experiment progress and total metric actions with our stakeholders.
Maturing an present OST
As you create every experiment thought, replace your OST to determine which answer thought you’re addressing with every experiment. The place you’ll be able to, annotate your experiment with a one-sentence clarification concerning the speculation you search to check.
Then, as your experiments conclude, floor the outcomes of every experiment in your OST as effectively. You should use a checkmark for validated hypotheses and an X for invalidated hypotheses, in order that stakeholders can rapidly see your progress with out entering into the main points of particular metric outcomes.
You should definitely broadcast updates to your stakeholders and managers each one to 2 weeks, relying on how rapidly your experiments resolve. By doing so, you show that your discovery efforts are bearing fruit, and that you simply’re not merely speaking to prospects for the sake of speaking to prospects. As a substitute, your discovery work is actively shifting enterprise metrics ahead.
On your first two or three broadcasts, you’ll need to pair these with dwell conferences as effectively. In these dwell conferences, you’ll need to recap that the purpose of the chance answer tree is to make sure that your product discovery efforts are shifting enterprise metrics, and that you simply’re delivery experiments quite than options.
In every assembly, give stakeholders the house to ask questions concerning the alternatives that you simply’ve prioritized and the answer concepts which can be on the desk. That method, they change into an lively participant within the product resolution making course of, the place they’ll contribute their concepts for potential buyer pains to deal with or potential answer concepts.
And, any time a stakeholder pushes for a selected function that doesn’t align with the metric or with a previously-identified alternative space, you’ll be able to gently however firmly clarify that their function is unlikely to maneuver prioritized enterprise outcomes.
When you’ve gained the belief of your stakeholders by these conferences, you’ll be able to ship updates asynchronously, to assist save on effort and time. That stated, make sure you set clear expectations that stakeholders can and will attain out to you if they’ve questions, considerations, recommendations, or new concepts so as to add to the OST.
Shifting to a brand new OST
After every quarter, you’ll must determine whether or not to proceed to remain on the identical OST or pivot to a brand new one. On the finish of the day, each OST will wind up with diminishing marginal returns.For instance, conversion charges can by no means exceed 100%.
Every time we determine that we are going to not get the best ROI by focusing on the preliminary metric, we have to determine a brand new metric to kick off a brand new OST. At that time, we have to restart your entire OST workflow that we mentioned in “establishing your first OST.”
Nevertheless, we will speed up the method a bit, since stakeholders ought to now be aware of the foundational ideas behind OSTs. It is best to solely want one to 2 conferences to align on the enterprise metric, on the chance house, and on the one alternative to concentrate on.
Upon getting the chance space to concentrate on, it’s best to be capable to run with options and experiments with out extra enter (supplied that you simply’ve constructed up sufficient belief over time with stakeholders, and that you simply’ve been addressing their questions as they arrive up).
We now know easy methods to share OSTs with stakeholders. However how precisely can we visually observe our progress as we run experiments? Under, we offer a template for easy methods to spin up a visible dashboard that gives related context and retains product discovery front-and-center.
Connecting OSTs to analytics
Alternative answer timber pair naturally effectively with analytics capabilities. In any case, we don’t know whether or not a given answer has made significant progress for the enterprise until we analyze its efficiency, and we received’t know whether or not we must always ship an answer or not until we will precisely craft an unbiased experiment.
For instance of the way you deliver your alternative answer tree to life by analytics, we’ll focus on how to take action utilizing Amplitude Notebooks.
Create an Amplitude Pocket book in your given alternative answer tree. This pocket book will function a centralized hub for info.
The very very first thing on the high of your pocket book must be the Key End result that you simply’re attempting to maneuver by your product discovery efforts. That method, we will see whether or not the important thing consequence has really been bettering over time primarily based on our experimentation efforts.
Proper after that, we then add our alternative answer tree diagram as a picture beneath the important thing consequence dashboard. We will create this diagram in any form of diagramming software, e.g., Miro.
Afterwards, we must always then checklist out every experiment that we’ve performed thus far. These experiments must be labeled with the given alternative that they’re focusing on within the alternative answer tree.
On the very finish of your pocket book, it’s best to embody an appendix that hyperlinks out to the entire buyer interviews and buyer analysis you performed to flesh out the chance answer tree. That method, you may have proof and proof of the alternatives that you simply determined to deal with.
Closing ideas
By utilizing alternative answer timber, we weave in steady product discovery all through all factors of the product planning course of. And, to assist stakeholders embrace product discovery, we must always use the chance answer tree framework.
We must always really feel empowered to tweak the method and the terminology to suit our particular contexts and wishes, so long as we hold these three ideas in thoughts:
- Product discovery should serve enterprise targets.
- The purpose of a product group is to iterate on options that may deal with buyer ache, quite than delivery options by some arbitrary deadline.
- We as a group should be capable to react to buyer learnings and suggestions in actual time, quite than ready till a quarterly or annual cadence.
Once we expose stakeholders and executives to alternative answer timber it can take a while for them to get used to it. However, as soon as we deliver them alongside on the journey, we’ll discover that they’ll be extra aligned with our product choices. Even higher, they change into lively collaborators and contributors, and so they’ll share extra concepts and buyer context than you may think.
If you happen to loved this publish, take a look at my others or go to Product Trainer for extra on product administration.