In this episode of the Subscription League, we explore the crucial balance between maximizing revenue and preserving subscriber satisfaction. Our guest, Vahe Baghdasaryan, a seasoned Growth Marketing Manager and 5-star Growth Marketing Mentor at CoinStats, offers invaluable insights into effective pricing strategies and monetization models for subscription apps.
Join us as Vahe highlights the importance of considering all aspects of monetization and emphasizes the significance of paywall optimization as just one piece of the puzzle.
You'll learn about:
1)the most commonly used pricing strategy
2)valuing features using Max-Diff Analysis
3)how to understand the willingness to pay using the drop principle
4)what to test when experimenting with pricing (subscription length, location, price)
Tune in for valuable tips and expert advice on optimizing your subscription app's monetization strategy.
Episode Topics at a Glance
- CoinStats’ approach to monetization
- Why understanding customers is important
- What is a banded model in subscription pricing
- How to value subscription features,
- How to measure the willingness to pay
- The drop principle - a commonly used product pricing strategies
- Price Experimentation - Big Testing for Big Results
- The impact of subscription length
- Location-based pricing
More about Vahe
Vahe is a highly accomplished growth product and marketing professional with a wealth of experience in consumer startups across various industries, including education and fintech. He currently holds the position of Senior Growth Marketing Manager at CoinStats, where he is responsible for driving subscription optimization, user retention, and activation and leads the lifecycle marketing team. He has spoken internationally on engagement strategies, user activation, and onboarding optimization strategies. He’s also a 5-star rating mentor on GrowthMentor. Vahe has recently released his on-demand courses on subscription optimization & pricing strategy. More information can be found here.
00:00 Welcome to the Subscription League
00:21 Welcoming Vahe Baghdasaryan
00:48 Vahe's transition from edtech to fintech
01:25 What have you learned from startup founders?
02:41 How has the FTX collapse impacted CoinStats?
03:53 Subscription optimization at CoinStats
10:04 Why pricing experimentation is important
15:02 What countries are driving most of CoinStats' revenue?
15:55 How do you set prices per region?
18:19 How to be careful about price discrimination
20:53 What is next for CoinStats?
23:00 Thoughts on the sunk cost fallacy
23:54 Where to learn more
25:00 Thank you for listening!
[00:00:21.560] - Olivier Destrebecq
Hello, everyone. I'm super excited to welcome Vahe Baghdasaryan to the show today. He's a Senior Growth Marketing Manager at CoinStats. Welcome to the show, Vahe.
[00:00:30.510] - Vahe Baghdasaryan
Hi, Olivier. Thanks for having me.
[00:00:32.260] - Olivier Destrebecq
You're welcome. I also have Jeff, co-founder at Purchasely with us today.
[00:00:36.580] - Jeff Grang
Hi, very pleased to be here, Olivier and Vahe.
[00:00:39.370] - Olivier Destrebecq
Yes, it's always awesome to have you. I like how we're switching co-host. It's like we're getting different questions, different angles. It's always nice.
[00:00:46.060] - Jeff Grang
Always a surprise.
[00:00:48.060] - Olivier Destrebecq
Exactly. Vahe, I was looking at your bio earlier and you've worked in edtech, now you're in fintech. To me, those seems like fairly different, but I'm sure there's a good explanation for your transition. Can you tell us more about that?
[00:01:01.040] - Vahe Baghdasaryan
Sure. There are fundamental principles in growth that is the same regardless of your industry, if you are in fintech, edtech or any other thing. From the edtech, I'm bringing with me all these fundamental principles and frameworks of growth. When you combine them with the specification of the fintech and crypto, it's more especially... It adds up in the end of day. Yeah, it works.
[00:01:24.920] - Olivier Destrebecq
Awesome. You've worked in growth in both of those areas, and today you're a mentor at GrowthMentor too. I'm sure it's a great feeling to help others answer entire question, but I'm also sure you've learned a lot by talking to startup founders. Can you tell us a little bit what you've learned from there?
[00:01:42.410] - Vahe Baghdasaryan
Yeah. I started at GrowthMentor more than a year ago, I guess. Many, many people come and share their struggles, and among them are startup founders. People who don't have growth teams are just getting started. They don't have a marketing team and they want to produce their growth initiatives by themselves. And then if they fund their product market, they're probably going to hire some marketing person, some growth team.
[00:02:07.380] - Vahe Baghdasaryan
During the discussions with the startup founders, I'm actually learning a lot from them about entrepreneurship, about starting startup, their struggles, how they manage everything by themselves. So actually it's getting beneficial both for me and both for them because they are getting answers from their questions and I'm getting a lot of information and knowledge about entrepreneurship and how startups work in a foot of a founder.
[00:02:36.220] - Olivier Destrebecq
I'm sure today you decide to create your own startup, you'll have plenty of knowledge right there.
[00:02:39.680] - Vahe Baghdasaryan
[00:02:42.440] - Olivier Destrebecq
Right now, you're a Senior Growth Marketing Manager at CoinStats, which is a pioneer in the crypto world. We're recording here, I don't know, a couple of months after the whole FTX debacle, but I'm sure that has impacted you guys. Can you tell us how?
[00:02:55.660] - Vahe Baghdasaryan
Yeah. I mean, FTX collapse was one of the most significant things that happened in crypto last year, and maybe the most significant of all the events that happened in crypto last year. It affected all the consumer sentiment of people who were in crypto. They were panicking and wanted to getting out of the crypto and starting to find other ways to invest, maybe.
[00:03:20.160] - Vahe Baghdasaryan
But the interesting thing was that we had this competitor who was somehow affiliated with FTX, and after the FTX collapse, all the people were accusing this competitor services, trying to found an alternative. We started launching some small campaigns to attract these users, onboard them through special promo codes and special customer support treatments, trying to find the growth of percentage even in the most dark times of crypto.
[00:03:51.000] - Olivier Destrebecq
Yes, there's always opportunities. Last time we've talked, you've mentioned that one of your big pet peeve when we talk about growth is that people tend to focus mostly on payroll optimization. There's plenty of other things to think about when you want to optimize your subscription. Can you walk us what you guys have done at CoinStats for this aspect?
[00:04:10.160] - Vahe Baghdasaryan
Sure. When you talk about a subscription optimization, a lot of people only talk about table optimization, but I believe that table optimization is only a small part of this whole monetization model and everything that you can do in monetization. Usually, it's not the first thing to start with.
[00:04:28.480] - Vahe Baghdasaryan
At CoinStats, when we started to change our pricing strategy and our optimizer monetization model, we didn't start with payroll optimization. We started with the most early stages of the monetization model to understand what are the use cases of our customers, and what are the problems that they are trying to solve with our product and also with our premium features.
[00:04:52.820] - Vahe Baghdasaryan
A lot of product teams miss this point. They end up diving monetization decisions without understanding these elements. And as a result, they are designing a system that is not connected with the customers, and they end up getting worse results or end up getting a model that is not very optimized.
[00:05:11.490] - Vahe Baghdasaryan
Usually, we believe that there are actually four parts of the monetization model that we put a lot of attention when we are starting. There were scaling how the model is going to scale, what features are we going to offer as a premium, what's going to be the price, and when we are going to charge for our premium features.
[00:05:32.650] - Vahe Baghdasaryan
When we talk about scaling, this is when... There are actually two ways to scale your monetization. The first is a continuous model. This is the model that most of the B2B companies are using. For example, Amplitude, Mixpanel, they start to scale on your pricing when you go over the limit that is based on your contract. For example, if you add up 10 million more events, the price will scale and you're going to pay more for them.
[00:06:00.440] - Vahe Baghdasaryan
There is the other model that is more common for B2C that is banded, and this is what we also end up using. It's like every premium package or every paid package that you offer that has some limitations. For example, if you are using our free plan at CoinStats, you are allowed to connect up to five portfolios to manage your funds. And if you decide to connect more, we have this premium plan that it's offered you up to unlimited amount of number of portfolios to connect.
[00:06:33.520] - Vahe Baghdasaryan
Or if you want to manage less than 1,000 transactions, you got to go with our free plan. But if you want to manage up to 1 million transactions, you got to go with our premium plan. So this is one of the important things to understand then.
[00:06:49.340] - Vahe Baghdasaryan
There is other things like what are the features that your customers are valuing the most? The interesting thing is that people tend to think that all the features are created equal, but it's not. And all your users valued the same feature in the same way. But in the end of the day, you have multiple use cases, and each use case value their features differently. For example, maybe there is this use case one that's value the feature X more, and there is this use case two that's valued the feature Y more. And you got to understand these differences before doing your packagings and starting moving forward.
[00:07:29.520] - Vahe Baghdasaryan
Also, the willingness to pay is different, and the willingness to pay for all your products and for all your features and all your use cases is different. This is the fundamentals to understand before moving forward and trying to go on.
[00:07:48.710] - Olivier Destrebecq
How do you dig in there? How do you go about understanding those?
[00:07:52.590] - Vahe Baghdasaryan
For feature benefit analysis, we were using this max tip principle of survey that is originally, if I'm not mistaken, presented by Free Forge. This helps you to understand what are the features that your customers value the most. You're going to use this, the most valued features, to show it on your payrolls, pricing pages.
[00:08:13.610] - Vahe Baghdasaryan
You can also take the least valuable things, this optimum valuable things and present it in your premium plan. Because if you are a company that is working with a premium model, you should understand what to put in a premium model and what to put in the premium plan. This help you to understand the position of your product or your features and use the best ones in promoting your premium plan and the least ones in your premium plan or other places.
[00:08:47.440] - Vahe Baghdasaryan
The other thing when comes to this, after when you understand this feature benefit analysis, it's also important to understand the willingness to pay off your customers. There are usually a few principles that do this. The most famous is the one waster drop that most of companies did and we also did.
[00:09:07.420] - Vahe Baghdasaryan
But there are also some companies but also doing some conjoint analysis or live testing that are more sophisticated. But I guess the go to choice for the most group professionals is this one where it's the drop principle, and it helps you to understand the optimal point to price your product so more users will convert and become paid users. And also it gives you some sense to understand how to discount your products.
[00:09:35.180] - Vahe Baghdasaryan
For example, it gives you some lower bound prices, and you can use that lower bound prices as a discount offer for your customers. And there is no limitation of how you can do this. For example, we did this in country-based. We did it on OS-based. For example, we did this for web users, we did this for iOS users, we did for the people who live in the US and Northern Europe and also people who live in Asia.
[00:10:04.440] - Olivier Destrebecq
You mentioned experimenting on pricing. Why is that so important?
[00:10:09.240] - Vahe Baghdasaryan
Because to put in simple words, it's going to help you to increase your monetization metrics. But actually, there are a lot of things that you can accomplish with pricing experimentations and monetization experimentations that can help you to improve your revenue significantly.
[00:10:27.700] - Vahe Baghdasaryan
There are many, many tiny things and also a big thing that a lot of companies, and they're getting started, or also if they're already doing well, but they're getting wrong about pricing and monetization. There are plenty of room, I believe, for everyone to optimize their monetizations. It's almost about everything, starting from the price that goes to visuals and the payroll series, et cetera. So optimizing them is going to help you to increase your revenue and also improve your monetization metrics.
[00:11:01.060] - Jeff Grang
Great. Can we speak a little bit about more precisely of the experiments you've been doing around pricing? Can you share us your feedbacks and experiences?
[00:11:09.720] - Vahe Baghdasaryan
Yeah, sure. When we started doing this pricing experiments, we decided to think they're not going to drive incremental results. So we end up doing a big testing.
[00:11:20.510] - Vahe Baghdasaryan
For example, when we started testing our pricing, we were experimenting 41.99 versus 100, 99.99, because most of the people when they start doing pricing experiments, they do, for example, 41.99 versus 44.99. The difference is really small. And if you even get a significant result, it's not going to help you improve your revenue significantly if you are not already... If you are not Tinder or Netflix, that even 0.5% change is going to help you to increase your revenue significantly.
[00:11:54.000] - Vahe Baghdasaryan
So we end up doing this big A-B testings 41.99 versus 119.99. And as a result of more than two months of testing, we end up getting twice more customers signing for 119.99 than 41.99.
[00:12:12.760] - Jeff Grang
[00:12:13.290] - Vahe Baghdasaryan
When you think about it, it's really crazy because the price difference is significant. There are more than three times difference in the price.
[00:12:23.640] - Jeff Grang
What would you say was the reason for that? Is that your product will be perceived in a different way, like better quality?
[00:12:31.760] - Vahe Baghdasaryan
Yeah, that's a good question. Actually, when you look at that, when a customer look at our competitors in the market, most of the competitors have more expensive packages than we do. For example, if we charge 119.99, most of our competitors charge at least three times more than we do.
[00:12:53.200] - Vahe Baghdasaryan
The other thing is when you price your product low, people tend to see these pricing in the spectrum. They see if your product is 41.99, they start to compare it with other product that it's in the spectrum. Unfortunately, these days, there are a lot of apps there that are trying to do their position themselves as cheap. But in the end of the day, they try to do scams. And maybe when people see the cheap prices, they tend to think that this could be not good or something like that.
[00:13:30.170] - Vahe Baghdasaryan
But I guess the main reason might be that most of the products in crypto are expensive. And when you see the product that is not as expensive as the others, you tend to be suspicious.
[00:13:42.060] - Jeff Grang
It's also maybe a population is more willing to pay a little more for a better quality of service.
[00:13:47.380] - Vahe Baghdasaryan
Yes. If you think about our customers, most of them have higher willingness to pay compared to other industries because these are folks who are doing investments, they're traders, and for a good services, they are ready to pay more.
[00:14:00.800] - Jeff Grang
Great. Have you made experiences of the subscription length? I mean, on the initial and basic subscription length from monthly to three months, for example. Did you do any experiments?
[00:14:10.020] - Vahe Baghdasaryan
No. Actually, we only experimented with two types of subscriptions, like monthly and annual, but we tended to push the annual hard to get users to sign up for annual. But for the most enthusiastic folks, we also offered lifetime subscriptions. That is five times our annual price. Some people in this trading, interested or more significantly, they tend to purchase lifetime subscription instead of monthly or annual.
[00:14:39.100] - Olivier Destrebecq
Did you have any other noteworthy findings through those experience? You found that you could raise the price considerably. Is there any other thing you found out?
[00:14:47.400] - Vahe Baghdasaryan
No, actually, we don't plan to raise our price at this moment, but maybe in the future we decide to do some other experiments. But at this moment, we are not doing any experimenting around the price of the product.
[00:15:01.080] - Olivier Destrebecq
Okay. In term of revenues, are there countries or region that are driving most of the revenue for CoinStats?
[00:15:08.280] - Vahe Baghdasaryan
Yeah. When we're changing the monetization model, we actually adopted the location-based model. But even before that, our mostly revenue came from United States, Canada, and the northern part of the Europe, and we saw there an opportunity to optimize pricing for the countries that have low purchasing power.
[00:15:28.980] - Vahe Baghdasaryan
So we end up adopting this location-based pricing and offer more comfortable prices for folks who lives in the countries that has low purchasing power, but also are enthusiastic in the crypto and there are some beginner traders and they want to start managing their funds. From there, we set up our prices for the regions and started doing other things.
[00:15:54.190] - Olivier Destrebecq
How do you set those prices per region, actually? Because A, there's so many. B, I'm curious what process you use to get to the prices.
[00:16:04.400] - Vahe Baghdasaryan
Yeah, sure. First of all, we divided location into these brackets, group them, cluster them, and divide into great brackets. We end up having five groups of locations, and we end up getting them based on our revenue analysis.
[00:16:20.660] - Vahe Baghdasaryan
For example, we saw that the most part of the revenue is coming from United States, Northern Europe, a few other countries. So this is a group A. The other part of revenue is coming from Eastern Europe, plus Russia, et cetera, et cetera. There's group two, et cetera. And after that, we started doing this one-way [inaudible 00:16:38] analysis to understand the willingness to pay per region and that's how we end up getting the location-based prices.
[00:16:46.800] - Olivier Destrebecq
It sounds like a fair amount of investment on your team's side to optimize this region, I guess. I'm sure there's some region that represent very small revenue. Is there big benefits in doing that price discrimination by region, I guess?
[00:17:02.180] - Vahe Baghdasaryan
Yeah. Actually, when we looked before, there are some countries where we're getting not a lot of revenue, and it's mostly the Eastern Europe countries because maybe our price... When we were adopted this location-based pricing and put a price there that it's more convenient people to use, they end up paying more. So if before we're getting the 1% of our revenue from Eastern Europe, now we're getting 2% of the revenue, 3% of the revenue from this year. So location-based pricing becomes heavy when you scale it.
[00:17:37.620] - Vahe Baghdasaryan
For example, if you tend to do a lot of price experimentation around location-based pricing, you need to have a big infrastructure to manage everything because there are several things going on the same time in different places and you need to have a proper infrastructure to manage all your experimentations and analyze them.
[00:17:57.860] - Vahe Baghdasaryan
But now we don't do a lot of things around this location-based pricing. We only sell it different prices per region. Maybe at some point in the future, we start to see that it's some location that has a potential and then we dive in there more and start to do more things to convert more people to our paid services.
[00:18:19.680] - Jeff Grang
Well, we've been discussing a lot about price testing, but the limit between price testing and price discrimination is very thin. Price discrimination is becoming a big thing. Tinder announced last year that they were moving out of this practice. We also know that the App Stores and Google Play stores are really looking at this practices. Even some local laws in Europe or in the US maybe are forbidding some of these practices or framing some of these practices. Going down that way, what should one be careful of?
[00:18:49.670] - Vahe Baghdasaryan
The first thing is I would advise to look at the local legislation where you want to start this price discrimination, so our pricing testing. For example, as much as I know in the EU, it's forbidden to do price discrimination. It's not if you can't price your product an X price in France and Y price in Germany, if it's the same thing. There are many, many other countries that they have their local legislation for price discrimination or pricing testing. First thing should be talk to the legal team and look at the legislation.
[00:19:28.480] - Vahe Baghdasaryan
The other thing is the most common thing that you can do is have just different prices on mobile and web. Many, many companies having more cheaper prices on web to help users to convert web and to avoid from this 15 or 30% commission fee that they pull charges. For example, if the product is like $50 on mobile, they offer 40 if you buy it on web.
[00:19:57.790] - Jeff Grang
Once again, that breaks one of the most common app store laws where the product should be sold at the same price no matter the mean or no matter where you're purchasing it from.
[00:20:08.600] - Vahe Baghdasaryan
Yeah, you're totally right. But I believe the line between what it's illegal and what it's tolerated to do is very, very thin. As long as you don't do it based on people's income or their gender or race, it's very thin. These are the things that are forbidden by all the laws. But if you skip this part and there's the line that it's very tiny between what is illegal and price discrimination and where is the opportunity to do price testings.
[00:20:37.840] - Olivier Destrebecq
I think at the end of the day, it's the difference between the law and the contract that you sign with Apple. I guess, your risk rejection from the App Store.
[00:20:46.000] - Vahe Baghdasaryan
[00:20:46.380] - Jeff Grang
And then navigating between these two can be a nightmare.
[00:20:49.940] - Olivier Destrebecq
They can make your life hell, but they won't send you to jail. You've already done a lot of optimization at CoinStats. What is next for you guys?
[00:20:59.040] - Vahe Baghdasaryan
We are currently exploring some behavioral science frameworks and cognitive biases, applications of cognitive biases that helps some companies to improve their metrics. There are some... And optimize their frameworks, optimize their metrics, [inaudible 00:21:16]. There are many, many cognitive biases that are involved in making a purchase. Some of them is the confirmation bias, sunk cost fallacy, and the endowment effect.
[00:21:28.930] - Olivier Destrebecq
Tell us what they are for people who don't know them.
[00:21:30.750] - Vahe Baghdasaryan
Yeah, sure. Actually, I'm just starting in this behavioral science thing, doing my research and exploring the things there. But as I feel, one of the best things that we are currently exploring is this confirmation bias. The confirmation bias is the tendency to search for, interpret, and recall any information, search for a social proof that confirms or supports their belief and supports their tendency to purchase.
[00:21:59.400] - Vahe Baghdasaryan
For example, if somebody wants to do $120 purchase, they tend to look for some social proofs or look at something that's going to say them, "You are doing right, go for it." And we are trying to have this confirmation bias element in our paywalls, in our landing pages to support our users.
[00:22:21.320] - Vahe Baghdasaryan
For example, we are putting reviews about our products from the key opinion leaders in crypto that there is extensive reviews that explain the benefits of the coins that we are putting, media mentions about CoinStats from a few of the most notable media outlets in the crypto, and experimenting with different things to help users to make sure that they are doing the right decision.
[00:22:47.780] - Vahe Baghdasaryan
In some cases, we are also offering them refunds. You have two weeks to apply for a refund. We are just getting started and we'll see the results in the future.
[00:23:01.140] - Olivier Destrebecq
Yeah. I'm curious because you mentioned the sunk cost fallacy, and this one is one of my favorite. I'm curious to see how to use it in subscription optimization, if you've already put some thought into it.
[00:23:12.120] - Vahe Baghdasaryan
No, actually, we didn't start it to using it. We are now more focused on this confirmation bias, but also doing a lot of research to understand how the sunk cost fallacy actually works. I saw a lot of folks doing this capitalizing on this, but we didn't start yet.
[00:23:30.720] - Olivier Destrebecq
For those who listened, the sunk cost fallacies, essentially, if you buy a ticket to a concert for 150 bucks and then the day of the concert, there's something else that you value more than going to the concert, you should just forego the cost of the concert and go to the thing you actually want to go because you already paid for it. So that money is gone. It's something that I found very interesting when I first heard about it at university.
[00:23:53.110] - Vahe Baghdasaryan
[00:23:53.120] - Olivier Destrebecq
You brought a lot of great value on the podcast. If people want to learn more about you and what you have to say about subscription optimization, where can they go?
[00:24:02.880] - Vahe Baghdasaryan
They can find me on Twitter and LinkedIn. I'm soon launching a subscription optimization pricing strategy course on Maven.
[00:24:11.380] - Olivier Destrebecq
[00:24:11.870] - Vahe Baghdasaryan
There is a waitlist open at this moment. I plan to start the course, maybe by the end of the February. It's going to be a one-week intensive course, seminars and office hours about subscription optimization and payroll strategy. It's not going to be only about payroll optimization. It's going to be defining your monetization model and all the down to creating lifecycle marketing campaigns that will support your subscription model.
[00:24:44.380] - Olivier Destrebecq
Awesome. Make sure to send us the link so we can put in the show notes.
[00:24:47.100] - Vahe Baghdasaryan
[00:24:48.430] - Olivier Destrebecq
Awesome. Well, thank you very much. It was so great to have you on the podcast. You had some great answers to our questions. Thanks again.
[00:24:55.090] - Vahe Baghdasaryan
Yeah, thanks for having me, guys. I hope I contributed something good to your audience.
[00:24:59.500] - Jeff Grang
[00:24:59.500] - Vahe Baghdasaryan
[00:25:00.410] - Olivier Destrebecq
On behalf of the Purchase team, thank you for listening to the Subscription League podcast. If you've enjoyed what you heard, leave us a five star review on iTunes or other audio platform. To find out more about Purchasely and how we can improve your subscription business, visit purchasely.com. Please hit subscribe in your podcast player and don't miss any future episodes. You can also listen to previous episodes at subscriptionleague.com. See you soon.