Summary
In this article, the aim is to explore a possible method of testing for Product Traction without building a full fledged offering at one go , thus, obviating the need to consume often constrained & expensive Technical resources.
The article proposes “Atomic experiments” as Risk mitigation technique & provides few examples as illustrations. The article makes the case that such Atomic experiments identifies risks early on & increase probability of creating quality durable products.
The theme of Atomic experiments is inspired from lean methods advocated for product building wherein build-measure-learn is the primary motto.
Introduction
A new Product being built from scratch has too much uncertainty baked in.
Hence, it is advisable to categorise the risks & demarcate ownership of identified categories.
Any new product being built has primary 5 risk categories to address before it can achieve PMF(Product Market Fit).
Value Risk - Does the Product provide enough value to the target customer and does the value appear attractive enough for the user to consider trying it out? Does the user derive enough value to pay for the Product?
Primary Ownership - Product Manager
Business Viability Risk - Does the current Product with its cost incurred & revenue opportunity meets the business goals of the organisation?
Primary Ownership - Product Manager
Distribution Risk - Can we reach out to the desired target segment , effectively communicate the value proposition & acquire users with low enough CAC to make the Product viable?
Primary Ownership - Marketing Manager
Technology/Feasibility Risk - Can we build the Product at minimum quality level & within required time with skills & resources in place?
Primary Ownership - Engineering Manager
Usability/ Design Risk - Can the end user figure out how to use the product & derive value out of it? For brand new Products requiring new behavioural set-up, risks of usability will be extremely high.
Primary Ownership - Design Manager
The above risks have sequential dependency as well.
Only when the the original Product Hypothesis is solid enough & value risks are mitigated, does the question of Business ,Distribution, Technology & Design risks come in picture.
It is apparent that without a product with strong enough core value proposition, any extra effort & resources put in distribution, business viability , tech or design aspects reduction will not yield the desired impact.
Alternately though , often quality products itself might not succeed due to distribution risks ( e.g. - targeting wrong segment, extremely high CAC) , Tech risks ( Slow buggy Product), Design Risks ( unclear call to action).
The example of TiVo is a testament to a great quality Product which could not realise its potential due to distribution risks.
TiVo was an awesome Product for the time , launched in 1997.
It had the capability to
Pause live TV,
Skip commercials,
Rewind Live TV,
Record shows automatically based on your viewing habits & many more.
The features were nothing sort of revolutionary for the time.
However, the Product GTM though was focused on the mass market i.e. anyone with a cable connection. The company, having raised enough funds, flooded TV commercial spots.
The outcome was that majority of mass market consumers did not care about ads or ability to pause Live TV. The inertia of existing behavior was too strong to adopt a new product which needs learning curve.
Very few consumers even understood the Product messaging due to sheer disinterest & technicality involved at times in using the Product .
Since the top of the funnel was the mass market & resources put in acquiring users were astronomical , post launch it turned out to be a commercial failure.
The most reasonable explanation for the failure is that - TiVO should have gone after the small segment of early adopters first - individual who despise ads & like having control over their TV schedule ( e.g. consultants,Executives or similar busy professionals). For brand new Products that require new behavioural set-up from target users , keen interest & motivation on part of users is pre-requisite due to inertia of existing behaviours involved.
Apart from easy solutions such as heavy incentives ( e.g. Cashbacks, free rides form Uber, Offers from Swiggy) , which is often unsustainable & leads to price war , the other durable way of establishing new behaviours is targeting the right segment.
Early adopters as a segment tend to be small , however at the same time quite influential & price “insensitive” . TiVo, by going after the mass market at one go , in a simple GTM language , filled with incentives that mass market understood, could not appeal to this influential segment.
Instead, the exclusivity factor of the Product could have been advertised to the early adopter segment only & then this segment’s influence would have likely worked as a organic customer acquisition channel for Early Majority- individuals who are influenced by credible others while making brand new purchases .
Side Note : - After Value risk , between Technology & Distribution risks, the risks of distribution is going up significantly compared to 20 years ago.
Dev & GTM in early 2000s
In the early days of 2000s, for building out a website, servers had to be bought, engineers had to be hired & only then Product could be built out. Broadband was still slow & hence websites had to make difficult trade-offs in features vs speed. It required immense capital upfront.
In parallel , customers could be reached out cheaply enough through search engine advertising or other similar avenues. Hence, Technology risks was higher compared to Distribution risks. The funnel at the top containing average new products launched in a year was relatively low.
Dev & GTM in 2022
With the advent of Cloud & No Code Tools, building MVPs is minor capital & effort expense. This has led to explosion in solutions available - both to B2C End customers & B2B businesses solutions as SAAS offerings. In addition, ad market has largely become duopoly with very high ad expenses. This implies very high Distribution risks compared to Technology risks.
The funnel at the top containing new Products launched in a year have exploded & will continue to go up over the next decade. The new constraints will primarily be centred around distribution risks.
Takeaways
Therefore, Product & Marketing must focus on GTM earlier in the process of building the Product than traditionally required.
If feasible , GTM assumptions must also be tested out on a smaller scale first to unearth new possibilities & nuances.
You can read on the topic in more detail here.
Implications
All the of the above point to the vital centre piece i.e. Hypothesis formation as the first piece of the puzzle to be solved with razor sharp focus. This is why the renowned Product Coaches such as Marty Cagan often encourage spending 95% of the time on the problem & the rest 5% on solutioning.
This ensures that as long as you have picked the right Problem statement to solve & are equipped with enough arsenal of novel insights - solutioning can always be iterated upon with newer hypothesis. While testing for hypothesis on a smaller scale, you are likely to unearth more insights to upgrade you existing Hypothesis.
But the opposite is not true. Often startups have to pivot precisely because solution iteration on earlier hypothesis does not yield the desired outcome and the team has figured out an alternate adjacent opportunity with the most potential as part of current trial & error.
Constraints
The roles & responsibilities of each team come out as :-
Product Team searches for Insights & builds Hypothesis
Marketing architects the GTM approach
Tech & Design brings ideas to reality
However, Technologies team is often the most overloaded of all & constantly have more to do than they can cover at the moment.
Is there an alternate method of building quality durable Products which solves for this major constraint?
Atomic Experiments - The Risk Mitigator
An atomic experiment is the smallest possible Product execution that tests for the Hypothesis assumptions, without building out the full fledged Product first.
Any hypothesis formed will have few assumptions embedded in them. After go to market, assumption risks are what might fail a potentially great idea.
Therefore, it is desirable to test the assumptions out first on a smaller scale without building the full fledged Product first.
This enables Teams to conduct multiple experiments in parallel , discover the potential risks & nuances early on and pick the Product category with the most potential for full fledged execution with Tech & Design involved.
Experiment Category 1
Minimal Design & No Tech involved
For example, let us consider below Hypothesis.
Insight - On social media channels, Products with Humans(Models) & High quality picture create the highest possible impression in mind of end customers.
Hypothesis 1- In a world full of mobile camera, End customers will prefer getting clicked by professional photographers. The resulting photographs can function as authentic Catalogue pics, Instagram wall posts & testimonials for the merchant.
Hypothesis 2 :- In case professional photographers turn out to be expensive, we will look for new age photographers who are young, inexperienced but well known on social media for their unique and novel approach to photography, preferably popular with GenZ.
Assumptions
Merchant will prefer sending photographers to end customers house/preferred location.
End customers are comfortable with photographers visiting them at their home.
End customer has the time & eagerness to put the effort in looking good for the occasion.
Potential customers find existing customers donning the seller’s offering(e.g. Apparel, Watch) as more credible & noteworthy.
For the merchant, the value derived from getting end customers high quality pics is lower than the costs incurred.
Conducting the Atomic Experiment
Since the Value risk still loom large over the potential product offering , involving Design & technology at this stage fully may not be desirable .
Instead , Product and Marketing can run smallest possible atomic experiments which test for the assumptions & also bring other possible nuances that team may not have visualised yet.
The smallest possible atomic experiment is done in ad-hoc manner with no digital product yet built out. While for target customer, this may not bring out the high quality UX. However, to begin with, even mild interest on part on target customers with lure of reimbursement of expenses ,for participating in such experiments ,is a strong traction signal on the core value proposition.
For the above hypothesis, The primary goal and execution steps are :-
Primary Goal - Make go/No go decision on the Hypothesis
Teams Involved - Marketing, Product, Design
Tools - Excel, Whatsapp, Phonecall, Mail
Execution Steps
Marketing identifies 5 early adopters or potential Merchants who fit the target merchant profile( e.g. Apparel, Accessories with Visual components).
Marketing reaches out to the respective merchants with promise of reimbursement of all expense for participating in the offering.
The interested merchants share the profile of the end customers they are interested in sending photographers to & criteria for selection of such customers.
Merchant reaches out to their customers with offer of getting clicked by professional photographers in Products.
The interested End customers city location is identified.
Relevant professional photographer is identified.
If feasible, for such experiments, Product/Marketing helps supervise the day of the photoshoot.
The photographer shares the pictures to the end merchant & to us.
We create a catalogue out of the pictures with help from design team.
We encourage the merchant to share the catalogue & the individual pictures on their respective social media channels.
We repeat this 4 more times or close it even before if Product team believes all relevant inputs available for making Go/No go decision.
If challenge of above ad-hoc approach fails, team should assemble a working prototype quickly using NoCode & attempt to get the users to use the product. Failure at this stage is expected & feedback will be used as an input for iteration.
While in the above example, Design team had minimal involvement, there might be certain experiments where small parts of the Product might need to be built out with tech team’s assistance to be able to conduct the experiment at minimal level of efficiency.
Experiment Category 2 - Minimal Design & Tech involved
For example, consider the below Insight & Hypothesis .
Insight - Gifting others is a major hassle & requires immense mental effort on part of “Gift Giver”. The Cash is often given as Gift Card due to sheer effort involved. There is an opportunity to replace such Commodity gifts with relevant unique Products to create delighter experiences for recipient .
Hypothesis 1 - Social Media has extremely high quality & variety of niche offerings & increasing. A solution which helps unearth appropriate unique items to “gift givers” will be significant value add.
Alternately, on an individual merchant level, the tool can be built out wherein the specific product is recommended to end customers based on inputs provided in the tool of the characteristics of “Gift Recipient”.
Hypothesis 2 - What makes for a good gift is based on “preferences of recipient of the gift” & “occasion” . Social proof from others as to what has worked in past will help decide the gift giver on categories to consider.
Hypothesis 3 - Certain individuals enjoy the experience of deciding on gifts & relish in the challenges of making the offering special. The gift giving effort can be outsourced to such individuals.
Ambitious Hypothesis 4 - A solution which enables gift giving will lead to a new behaviour of impulse gifting ( similar to impulse food ordering as a second order effect due to availability of Swiggy/Amazon). In this cases, users might start gifting to individuals they have not done before - e.g. to your siblings, cousins, close friends., Neighbours just because the Product makes it extremely convenient.
Assumptions
Social Media Merchants consider their SKU category as worthy of Gifts & consider Gift category as attractive enough for their business.
Social Media merchants will offer the tool to their end customers.
“Gift Givers” will find the recommended options valuable enough.
Our recommendation algorithm is efficient enough to surface relevant unique items from Social media Sellers.
Social Media sellers will be willing to be aggregated on our platform so as to ensure their discovery to “Gift Givers”.
“Gift Givers” are keen enough to delight and make the gift memorable for the “Gift Recipients”.
Conducting the Atomic Experiment
For the above hypothesis, The primary goal and execution steps are :-
Primary Goal - Make go/No go decision on the Hypothesis
Teams Involved - Marketing, Product, Design
Tools - Excel, Whatsapp, Phonecall, Mail
Execution Steps
Marketing identifies 5 willing users who have a need to gift something to someone.
Product takes all the below requirements from the user-
Receiver’s profile ( Gender, Age range, Professional, City)
Receiver’s like/dislikes & extra curriculum interests
Occasion of gifting & date
Giver’s Profile - Gifting purpose, Budget , if applicable, preferences etc.
Product & Marketing identify 2 individuals who like the exercise of gifting & are willing to help us figure out Product Categories to gift( Apparel, Accessory, Book etc.) based on input data.
Product with help from identified “Gifting Experts” recommends categories to the end users on survey/some other tool.
End user rejects/select the categories.
Post category finalisation, Product team recommends Specific item SKUs.
End user selects/rejects the SKU.
The goal of this exercise to figure stand parameters that applies in gifting for this particular combination of Giver’s profile, Receiver’s Profile , their relationship & the occasion itself.
The insights generated will form the algorithm that will be developed subsequently.
We repeat this 4 more times or close it before if Product team has all relevant inputs available for making Go/No go decision.
The interaction steps 3 & 4 above might require some tools to be developed with Tech team involved to ensure efficiency.
Experiment Category 3 - Minimal Design & Tech involved
For example, consider the below Insight & Hypothesis .
Insight - End customers really like a certain Product immediately when viewed on social media . However, need for purchase might be some time away and the customer may not be committed for purchase right now(e.g. Festival, Event to attend 6 months later). The interest on part of End customers & upcoming occasion for purchase is not getting captured currently.
Hypothesis 1 - Information such as current interest in the Product & when the purchase might happen will be extremely valuable for the social media seller. Currently end customer track it passively in Bookmark/Save options in Instagram. A journey wherein Merchant proactively reaches out at the right time with the right Product will lead to more purchases.
Assumptions
Merchant will be willing to attach our tool’s link in the Instagram posts/profile/send it on WA to end customers.
End customer will be able to discover the offering.
End customer will be eager enough to try out the solution & provide relevant info for reminders to be sent.
Merchant will have similar/relevant SKUs during the occasion formed as part of the reminder.
Conducting the Atomic Experiment
We will need to build the minimum viable product for the test to be done for at 5-10 merchants.
Product Team , with tech team, builds out a simple “Interest Capture” tool only , configurable for one merchant at a time .
Marketing identifies 5 early adopters or potential Merchants who fit the target merchant profile.
We take Profile data of 1 merchant & hard code the profile details in the tool.
We request Merchant to attach the “Interest capture” tool link in the next 5 posts & stories of their SKUs.
We measure the effectiveness of the tool.
We repeat the process across 4 more sellers.
In summary, there will probably be no stand format as to how to conduct the Atomic experiments. It requires speed and flexibility so as to be able to mould itself as per the Hypothesis. The adoption of such practices will likely also solve for certain inevitable biases , within team & across functions, that leads to incessant debates with no conclusion.