Matt Wright: Let’s get started with a basic question. With the user in mind. How does did users historically buy and use ink? What are some of the classic struggles that a user has from an ink perspective? 

Sovrin Tolia: Yeah, fantastic question. So let’s think about the customer. What is the role of ink in their lives? Ink is a means to an end. The customer is looking to delight, create, learn, manage, transact with a print outcome. The role of ink is to fulfill that.

Historically, one of the biggest pain points with ink is running out at a bad time. It’s 7 a. m. Kids need to turn in homework. You need to print it. You fire the print button. Printer is out of ink. Worst situation on a Monday morning. 

Ink can also be expensive for a lot of customers. While the delight of printing is very real, the cost associated with it is prohibitive. 

Thirdly, is just the hassle and stress of having one more thing to buy and manage. Which cartridges do I buy? where do I buy it from? So it’s one more thing in their already busy lifetime. 

So that’s why listening to our customers, we innovated on the HP Instant Ink service to solve these problems, we attack the problem of convenience by promising never running out. We’ve changed the business model, so instead of buying ink, you sign up for a monthly paid plan that provides up to 50 percent savings on ink. We’ve taken the hassle out of the process,  so that customers never have to worry about ink again .

Matt Wright: Super fascinating. If I just dig into that a little. How were these pains identified and did HP then shape them into key value propositions for the service.

Sovrin Tolia: The business and the product value proposition has been built generation to generation. We’ve done a lot of customer interviews. We went through a lot of reviews provided by customers on web forums, Reddit, and Trustpilot. We’ve partnered with Conversion, you guys, to do thousands of experiments, to learn about what the customers are telling us, what kind of value prop resonates and how to build a value prop that meets the customer needs.

When you combine these methods together, it gives us a powerful sense of customer struggles. Then testing methods give us a way to validate some of our solutions to those at scale. So those methods have always helped us listen to customer’s problems, come up with a hypothesis, test it and scale it. So its continuous cycles of test, refine, test, refine, scale. 

Background on Instant Ink

Matt Wright: How long has instant Ink been around?

Sovrin Tolia: We’ve been in market for about 10 years now. We have 10 plus million subscribers. We are a very healthy, growing subscription business. We are one of the best case examples of a true monetized paid subscription service in the consumable space operating in 38 markets.

We’ve essentially transformed a desktop printing experience into a fully cloud connected printing experience with no print interruptions. Historically, printers were a desktop device connected by a USB cord to a PC. 

We changed the paradigm to make printers first class citizens of an IoT (Internet of Things) ecosystem. This closed loop cloud connected IOT ecosystems, the printer tells us it’s ongong ink status and our fulfillment algorithms decide when is the best time to send a replacement cartridge to ensure the printer doesn’t run out of ink. Then internally we innovated on ourfull supply chain network to make sure the ink gets to the customer on time.

Instant Ink GTM

Matt Wright: How did HP introduce Instant Ink and then start trying to grow it? 

Sovrin Tolia: This is typical of not just back then, but even today at HP. It starts with what the customers are telling us and identifying what problems to solve. Then always starting with a very quick pilot.

I remember when we first piloted Instant Ink, we had a team that was selling the service to Best Buy customers by standing in store talking customers through what it would take for them to sign up. They would agree to sign up. We capture their information. And then somebody in the office would manually ship them cartridges. That’s how we validated the value prop.

Typically the approach we take is user testing for various iterative concepts and solutions first to get early feedback. Then we do an in-market pilot on various offers, positioning solutions, etc. on a small isolated scale. Then we start building that exposure towards validity for scale. Once we have the the solutions proven at scale, we expand. 

We started with a very small pilot in the US and then expanded nationwide. Then we expanded to North America and Europe in three phases over a four year period. We went from zero to 38 markets, by iteratively learning along the way. 

Matt Wright: What were some difficulties getting someone to sign up for Instant Ink?

Sovrin Tolia: I think as much as printing is a delightful experience, having to worry about printing is not something at the top of a customer’s mind. 

Printing behavior has been ingrained in the market for 20 plus years. Consumers had a set way they think about ink, they had a set way to think about where they purchase it. 

So one of the biggest challenges we had was disrupting this ingrained consumer behavior. It was a big paradigm shift, ifyou’ve done this for 20 years, you don’t necessarily have to live through the pain anymore. 

The second big element was time. When customers think about buying ink or replenishment for their printers, they don’t necessarily think [about the process] too much. We do not have an infinite amount of time to sit in front of customers and explain what the service is about. Especially in a noisy retail environment, customers have a very limited attention span.

A lot of innovation was with the design intent, if you had five seconds to convince the customer to sign up for the service. How would you do it? How do you bring the value and benefits in a way that they can digest quickly, make a decision, and then realize the benefits. 

Matt Wright: The decision to onboard people in the printer flow was on the roadmap? Or is there something that led to “hey, this is something we should try”?

Sovrin Tolia: So printer setup is one of our most successful onboarding channels. The hypothesis was, look a customer who buys a printer is bombarded with a lot of offers to buy additional supplies. Buy this, buy that, buy a warranty, etc.

So to this philosophy of building an IOT system we thought we could frame Intant Ink as a feature of a printer. It’s cloud connected and it replenishes automatically. So we created this concept that positioned Instant Ink as a feature of the printer that is available to a user if they decided to turn it on during printer setup.

That changed the mindset of the customers. It’s no longer an offer or a sales pitch. It’s more about my printer is capable of doing this, do I want it or not? 

The hypothesis was strong, but by no means did we know how successful it would be and it still took years of refinement throughexperimentation of the printer set up onboarding flow. How do you talk about the service, what kind of offers you want to bring, what kind of imagery that goes with it, should you talk about the cost of supplies that’s in there?

It took the full spectrum of experimentation, messaging, and offered value proposition benefits for eight plus years in printer setup flow. [All of that combined] has led to our success. 

Matt Wright: Eight years ago is when we got the opportunity to be involved with this and go along on the journey. I think HP Instant Ink is this amazing case study of product led growth. How did you go about growing through the product itself? 

Sovrin Tolia: When we started with printer setup, we didn’t know how to position it. 

The first approach with Conversion was essentially an advertisement. Once printer setup is done, we would ask if the customerwould like to have Instant Ink, [positioning it as] It saves you money, 50%, [and you] never run out. [From doing this] we got a certain take rate.

But, the customer feedback was, why are you bombarding me with offers during printer setup?

The next evolution was look, we need to reduce the barriers to enrollment. Printer setup is not the time when somebody really wants to make a commitment decision. So let’s take commitment out of the process. So we innovated around a free trial. For the first three months on a brand new printer device we offered a complete free trial on us. 700 pages per month, more than most home printers can ever print. We really unlocked the freedom to print with no commitment trial that customers could always cancel.

How did we do this? We did this through experimentation. Where we said, “hey, we need to reduce friction, lets test the concept of a free trial”. We went into OOBE (Out of Box Experience) and did a true AB test with a free trial vs. no free trial, and we started seeing conversion uplifts that were triple digits gains.

Once the free trial was successful, we said, okay how can we reduce friction more? Let’s test a different kind of product led growth model, which is freemium,“hey, you don’t have to pay me a thing if you don’t print”.

The commonality between both of these is we believe our value prop is real, it’s foundational and strong. The philosophy of product led growth is you consume our product. In this case, you consume instant ink, we send you cartridges, you love it so much that you will stay forever.

Both these business model innovations happened because of our strong experimentation approach in partnership with you guys at Conversion to test, learn, test, refine, test, learn, refine. 

Again, the free trial and freemium were not hits from day one. There wasa ton of learnings along the way to refine from a concept idea to a solution that scales to hundreds and thousands of millions of subscribers. Its a long journey and hundreds of experiments.

Stephane Bardin: I want to add something interesting here about how the progression of experimentation and modifying the business model, to the point where now the business model is improving the sales of printers, because these printers come with free ink and this service.

So not only have we added a service to improve your printing experience, but now it’s full circle. We’re now influencing consumer perceptions when buying a printer,“HP Printers just makes my life so easy because it comes with that service.” 

Sovrin Tolia: That’s so right on. We drive stickiness to the service. Stickiness drives loyalty and delight with printing. Which leads to a user that is super satisfied with their printer.

The service creates a lot network effects around this device ecosystem. Supplies helps printing, printing helps printers, and the customers go and buy those printers. So that loop becomes a self reinforcing loop for us because the experience is so seamless for our customers.

Growth Levers

Matt Wright: In terms of the big growth bets or levers that you’re pulling, how does HP go about identifying what those might be? 

Sovrin Tolia: With Conversion we’ve established a very democratic experimentation process. So think about all the creative functions within the organization, marketing, design, product growth teams and retention teams. All of them come up with ideas and we’ve enabled an experimentation process where every idea gets evaluated for its strength of the hypothesis, and then gets it gets tested.

The other big lever for us on driving new growth vectors is a very strong set of analytical insights about that. Our subscription business is powered by very strong analytics about what every customer is doing with respect to their service and printing. So we use a lot of machine learning algorithms and predictive models to understand where different customers are in their process and use those insights to drive growth vectors.

I’ll give you a simple example. We might have an idea to create a 75 page plan? Would it drive growth and retention? This idea might be the results from insights generated through analytics, what customers are using, and the plan movements we are observing. If the evidence is strong we will go to Conversion and work with you to refine the idea and test a 75 page plan.

Using an initial test we will try to get a preliminary read of the impact of a 75 page plan on acquisition and retention. If the signals are promising then we start a broader scale plan. Let’s experiment at statistical significance in multiple markets, and then validate the customer lifetime value which is the end metric we target against. Then start scaling. 

So I think the scaling process, the ideation process, the growth process is a blend of democratic ideas from everybody in the company who’s associated with the service to very bottom up insights about what our customers are doing.

Alex Mason: I think the important thing to draw on here is that it wasn’t just one test that we did to get to this very specific answer. It wasn’t just like we turned this one thing on and we’re like, yes, we have the answer. It was a multitude of tests that gave us a much better direction as to where to go that guided us towards the right way.

Matt Wright: I think what makes the team around instant ink really unique in my opinion is theculture you’re talking about, which is that lots of people are constantly contributing ideas but balanced with a team mindset where everyone is bought in that ideas need to be tested, validated, and refined. How did that come to be? How was that created? 

Sovrin Tolia: A colleague of mine Gretchen Gary was the first to bring AB testing into the program. AB testing in the printing space was a very new concept. Nobody was doing it, we were selling boxes. So we had to drive a paradigm shift which Gretchen pioneered. 

So it started with establishing belief with senior leadership around the benefits of AB testing.

We started very simply. Let’s establish a process. Let’s do tests. Let’s see the benefits. And over time, as tests started to prove out, as we showed business benefit, as we drove conversion rates, enrollees, customer lifetime values the belief in the organization started to form around every idea needs to be tested.

We made large bets, tested in bits but not in totality. This led to some learnings around it where we said, “huh, if only we had tested this before, we could have executed more effectively”. 

So a lot of on introspection reinforced the need for multiple levels of testing. Starting with a basic resonance check, “Does it even stick on the wall?” And then a statistical significant test, “Does it resonate with a large group of customers?” And then a multimarket test, “Will it stick across all the world?” 

So it took us time. But it started with establishing an A B testing process, showing short results to the organization, and then over time instituting that culture of saying every idea is a good idea, but it needs to be tested. 

Matt Wright: It’s a humbling thing, right? Like I think you always think you have great ideas but there’s so many unintended effects. You’ve mentioned different business models and different markets. HP has tried different touchpoints too. What did that look like?

Sovrin Tolia: For us, a lot of evolution happened on our digital touchpoints and journeys. Both from expanding the touchpoints to improving the touchpoints in isolation. We started with a flagship printer setup. Over time, we went into smart apps. Which we call smart digital experiences.

HP has a smart app where millions of customers engage every day. The monthly active users are phenomenal. We have good signals about printing usages that enables us to do segmented offers. If you print 20 pages a month over the last 12 months and print a lot of photos we can recommend the relevant Instant Ink plan for you.

We have a flagship experience on our web property called instantink.com. All the worldwide marketing campaigns, onboarding traffic comes to instantink.com so this is now also an important enrollment funnel.

We do a lot of optimizations there. From landing page optimizations on what we communicate and how we communicate to the full blown end to end enrollment funnel. How to reduce steps, how to reduce time.

So across all of our touchpoints we are constantly doing a blend of messaging, marketing testing coupled with enrollment flow optimization.

Matt Wright: There’s some great stories about unexpected learning testing across these touchpoints. Alex can you share a some insights we’ve learned by trying to use the free trial lever more broadly? 

Alex Mason: Yeah. We had introduced the free trial and I think it’s still offered quite widely within the program. But we learned its effectiveness does depend on the actual touch point and where are the customers at in their journey. And in some cases can have inverse effects.

When we first tried testing the free trial on instankink.com we actually saw a huge decrease in the number of people actually considering and enrolling in Instant Ink. At the time this was a very surprising result because it was such a knock out of the park experience for OOBE (out of the box experience). But when we went to this other touch point with it, it was a total flop. 

We learned that at that stage in the journey consumers were not ready to make a decision about a free trial. And if we didn’t have the program, it would have been a very tough thing to to swallow that quarter with the decreases that we meaured. 

Sovrin Tolia: Yeah. And to build on that more broadly we’ve learned through the experimentation a one size fits all approach does not work in our cast.

So we always have to validate that if something works during printer setup, does it work to in other touchpoints like the HP Smart app. To Alex’s point, these customers are at different points in their thinking life cycle and so not everything resonates equally as well across the board.

On the free trial, I’ll give you one more example of the evolution. So like I said, we started with printer setup. We would offer X months of free trial. We tried probably 40 to 50 experiments on how to talk about the free trial because free trial comes with a psychological barrier of, oh, there’s a catch.

At the end of the free trial, I will forget, and then you will keep charging me forever, right? So while free trial reduces the barrier, it also raises some questions in front of the customer. So we had to battle all the psychology elements that our customers were grappling with to position and communicate effectively.

So we’ve tested things like whether the elasticity between three, four, five, six months, That’s one vector of testing. We’ve tested different positionings. We’ve gone all the way to what is the difference between a commitment based trial and a non committed trial. 

Meaning there’s a free trial where you need to fully enroll. You need to provide your shipping, your billing, etc. Then a non committed free trial truly was, where you just give me your shipping address. I will send you our Ink and then you have three months to try out the service. If you like us, give your payment information and we’ll be on our merry way.

So that whole spectrum of how much a customer needs to commit towards a free trial. We’ve explored that spectrum in its entirety. To reduce as many barriers and as many as much friction as possible. One example where it works out is if you’re doing printer setup, you love the free trial, but it’s like, oh, I don’t know where my credit card is.

Or Oh my God, my husband has my credit card, I don’t have the numbers. We want to reduce those barriers, right? So we move to a non commitment free trial, making payment methods optional, etc. All of these tactics of experimentation on a free trial was powered by AB testing. All of these were where we together would ideate around solutions beased on what customers tell us and identifying where the friction points are. We would go and say, “Hey, how about if we test a free trial where the payment method is optional?” That leads to the test idea. Then we go and design an experiment. We measure the onboarding rates for the free trial. We measure the free trial retention rates. If we like what we see. We scale it.

This is a great example of how you can use experimentation effectively to actually find was of reducing friction. 

HP’s Printing Flywheel

Matt Wright: Definitely tons of refinement and nuance it feels like all the time. Maybe that brings us to a nice point to talk about what comes next to continue to build and refine the service. 

Stephane Bardin: So here we are, we have consumers who are interested in printing. They buy a great printer, they buy ink that is technologically advanced, but then we leave the rest for you to figure out on your own. 

What you need to print is three things. You need a printer, you need ink, and you need paper. The idea of the paper add on service came about because of a belief that we need to complete a printing flywheel.

The introduction of the paper add on service was a natural completion of the printing experience increasing the value to consumers. You run out of paper just as much as you run out of ink, right?

I would say even more of a fustration because you can really only use ink for printing, but paper, the world is your oyster. You want to have an airplane paper war, go for it. You want to, doodle, go for it. You want to do arts and crafts, go for it. So paper is really used in much more varied ways than ink. So having a service that allows you to not have to worry about it we thought could definitely be enticing for consumer.

“Don’t worry paper. We’ll take care of it.”

The last thing which goes to the flywheel is how does the consumer judge a printer?

They can look in the box, but really not going to see anything. We can say it prints 100,000 sheets per minute but that’s really just a meaningless performance metric.

At the end of the day is when you get that great printout. That’s when you start feeling the value of ink and the printer. So that completes the flywheel where the consumer feels completely taken care of. I’ve got the best of everything, and I don’t have to worry about it. 

Matt Wright: I find the paper add on fascinating. Just like Instant Ink, getting a paper subscription is a very different type of model for consumers. How did HP test and validate that this is a service they should offer? 

Stephane Bardin: We start by talking with our existing customer base and asking how they felt about subscriptions? What role do they play in your life? When does a subscription model make sense? When does it not?

Then we do all kinds of surveys. The problem with surveys is at the end of the day you can ask the customer, ‘Hey, do you want a Ferrari?’ The customer would say ‘Sign me up’. But now, is the customer really going to pay for a Ferrari? That’s a different story. 

So what made a difference in our confidence levels that this was a good service to offer was working with Converion to do AB testing using a painted door approach where people are actually signing up for the paper service, even though it didn’t really exist at the time and measuring willingness to pay at different pricing tiers. 

Running tests where people are actually putting money on the line is so much more powerful to validate an idea. That really unlocked and started the process for us. We’ve got something here that consumers love. Let’s go for it. 

Matt Wright: How do you actually go about like even designing an experiment that tries to give you that type of insight?

Alex Mason: It’s tricky because you often have to pull apart the hypothesis and the actual execution of it. We worked closely with Stefan and team. And it wasn’t just one go. It was multiple goes over multiple months before we got to a place where It would work with customers.

We had the insights into sign-up rates and willingness to pay at different price points, but it was also a matter of integrating it with the existing Instant Ink offering. The core product is ink, and when you introduce this new concept it can be a friction point.  So it did take a number of tries to actually introduce it. 

Painted doors are a powerful tool for a lot of very early exploratory work to gauge user interest. Follow-up validation tests making sure that we’re not actually, harming customer sentiment or affecting their experience especially when adding on to an existing product or service. 

This approach really helps to inform these larger strategic decisions for the business.

Matt Wright: Was there any challenges you faced once you validated that the paper subscription was a good idea in terms of taking it to market and making it work with the existing offering around Instant Ink. 

Stephane Bardin: Ot’s one thing to validate the idea, but then it’s another question at what point is best to communicate it in the customer journey. 

Is it an add on? Is it a complete package? Do you do it at the very end of the process or do you do it earlier? 

Now that we know that there is a demand for this service, we started with a group of existing Instant Ink subscribers and offer them to test out the paper service subscription add-on for 6 months. They actually got to experience the service. To make that happen I personally got to ship hundreds of reams each and every day packaging them into boxes, putting the FedEx labels and making sure they arrive on time to these test users.

This allowed up to validate and get feedback on the service over a longer period of time. We actively engaged with these tests user and constantly asking them questions. How was itdelivered? Was there damage? What was ther perception of the experience? What delighted them? What frustrated them?

We tested ever single part of the experience with consumers through that six month pilot to really refine the service. That was alot of iteration and  a lot of learning.

We then continued to optimize with AB testing. Back to a point that was made earlier, it’s one thing to optimize the OOBE onboarding flow , but then you need to start opening up to all the other touch points. Where, how, and when do you introduce the paper add on service in these flows?

We don’t want one service to overshadow the other. We want the services to compliment each other.

Sovrin Tolia: To emphasize Stephan’s point. This took instant paper service, took experimentation to the next level. Not only was a charter to grow instant paper subscriber base, but also grow in a way to maybe even protect the instant ink, or even grow instant ink.

Both boats needed to grow together. The experimentation team here, Conversion, HP together, we’ve mastered that art of how to look at tests from multiple different lenses and business impact.

Tips & Takeaways

Matt Wright: Is there any tips, takeaways of how to do that? 

Stephane Bardin: Never stop exploring. You need to really turn every stone, and Essentially check all your assumptions.

Every touch point is different, one size fits all doesn’t necessarily always work. 

This week I got an AB test that proved not good for the paper add-on service, but great for Instant Ink okay. So what’s where we dig into the insights and explore an iterative experiment. Let’s just go a little to the side of that experiment and just see what created that hit and now let’s dig into it and iterate from it. It’s really that curiosity and that iterative process that just helps us move things forward. 

Alex Mason: I would say that even after seven years of working closely with the team, I am still surprised almost every single week when we are running tests, because a lot of the times there’s these very minute details, especially in touch points that cause completely opposite effects of what we would have expected and what we’ve seen elsewhere.

And so to the theme earlier, one size definitely does not fit all. So it does require a lot of steps along the way, even just to launch in one market, let alone worldwide.

Sovrin Tolia: I’ll add to that a little bit. I think to me, the success of the experimentation program is the end to end nature of it.

We don’t go and plan individual tests. We think epics and themes. We start with a business problem, we talk about all the hurdles to the problem, potential hypothesis, and it’s a multi month, hundreds of tests kind of relationship with Conversiont’s where there’s almost a sense of continuity around the whole space.

Number one. It’s not a per test effort. But hey, look, we need to solve this for low usage customers, or we need to solve for instant paper growing double digits by protecting instant ink, or we need to solve for high usage customers. So we started that thinking process from strategy, ideation, and test into, I think that’s process that we have together is fabulous, right?

Number two, I think both of us together have done a great job. of making each person on the test team very knowledgeable about our customers, right? So through research, through insights, through A B testing, we try to share as much as possible, pretty much everything we know about the customer. So everybody’s on the same page and everybody knows what problems we are solving .I think to me, the key success factor to drive this level of sophisticated experimentation programs is not just about one test. It’s about moving the business forward in the next direction grounded in an end to end relationship, and then making sure all the involved leads have a common understanding of the customers. 

That has been fabulous, which I think in my experience, this is really a differentiator for HP and Conversion together.

Alex Mason: If I can add onto that as well, it’s like a lot of the time when we are working together on these things, we’re focusing on the customer problem rather than trying to push a specific business objective.

When you are a lot more customer centric you’re much more likely to find that elasticity or have big bet actually hit versus ‘Hey, we’ve got to launch this thing, how can we make it work?’ That constant orientation and in the eyes of the customer and using them to guide the decisions that we’re making has been one of the most powerful things. 

Thank You

Matt Wright: I want to take the big thank you to all three of you.

Thank you. Thank you so much for giving us the time to do this. Very interesting to go through all this and greatly appreciated. Awesome to reflect on that.

Sovrin Tolia: Thank you for the partnership. Phenomenal. 

Stephane Bardin: Thank you so much. Fantastic. 

Alex Mason: Thanks everyone.