Tag Archives: engagement

Announcing the next Retention Deep Dive, Growth Series, and something new

Over the last few years, I’ve worked with Brian Balfour (CEO Reforge, formerly VP Growth @ HubSpot) as a Growth mentor and contributor to the Reforge programs. These are part-time programs (no need to take time off work) specifically designed for experienced Product Managers, Marketers, Engineers, and UX/Designers in both B2B and B2C companies.

Today, Reforge announced their three upcoming programs this fall:

1) The Retention + Engagement Deep Dive program. I worked closely with Brian developing this program, which looks at every aspect of retention including activation, engagement, resurrection, and churn.

2) The Growth Models Deep Dive program. This is a new, detailed examination of a key growth topic Brian and I developed this year with Kevin Kwok.

3) The Growth Series program. This is Reforge’s flagship program that provides an overview of the key topics in growth that’s been 100% revamped to reflect today’s growth challenges.

Apply to Reforge (Takes ~5 minutes)

Each Reforge program runs from September 24th through November 16th. Seats always fill up fast, and I’m excited to be involved. I’ll also be doing some speaking and Q&A during the events.

Besides Brian, Kevin and myself, other hosts include Andrew Chen (General Partner @ Andreessen Horowitz), Shaun Clowes (VP Growth @ Metromile, former Head of Growth @ Atlassian), Dan Hockenmaier (former Director of Growth @ Thumbtack), Heidi Gibson (Sr. Director of Product Management @ GoDaddy), and Yuriy Timen (Head of Growth @ Grammarly).

About the Reforge Programs
These are all invite-only, part-time programs that last 8 weeks. Each program requires a time commitment of 4 – 8 hours per week. They’re designed for Product Managers, Marketers, Engineers, and UX/Designers in both B2B and B2C companies looking to accelerate growth in their companies and in their careers by developing a systematic approach to thinking about, acting on, and solving growth problems.

In addition to the course material, we’ll also hear from leaders in the industry through interviews, live talks, and workshops, including:

Fareed Mosavat, Growth @ Slack
Ken Rudin, Head of Growth, Search @ Google
Brian Rothenberg, VP Growth and Marketing @ Eventbrite
Ravi Mehta, Product Director @ Facebook
Mike Duboe, Head of Growth @ Stitch Fix
Josh Lu, Sr. Director, PM @ Zynga
Guillaume Cabane, VP Growth @ Drift
Matt Plotke, Head of Growth @ Stripe
Joanna Lord, CMO @ ClassPass
Gina Gotthilf, ex-Growth Lead @ Duolingo
Elena Verna, SVP Growth @ MalwareBytes
Kieran Flanagan, VP Growth/Marketing @ HubSpot
Naomi Pilosof Ionita, Partner @ Menlo Ventures
Nick Soman, ex-Growth Product Lead @ Gusto
Nate Moch, VP Growth @ Zillow
Simon Tisminezky, Head of Growth @ Ipsy
Steve Dupree, Former VP Marketing @ SoFi
See the full list here

Here’s some more detail about each program below:

About the Retention + Engagement Deep Dive
Retention + Engagement Deep Dive zooms in on one of the most important sub-topics of growth.

Retention and engagement separates those companies in the top 1% of their category. Every improvement in retention improves acquisition, monetization, and virality. But moving the needle on retention is hard.

This program takes a microscope to every aspect of retention, including:

  • Properly define, measure, segment, and analyze your retention
  • Find and quantify the three moments every new user goes through to create a long-term retained user
  • Construct a high performing activation flow from the ground up using detailed strategies across product, notifications, incentives, and more
  • Layer your engagement strategies to build a compounding growth machine at your company
  • Articulating retention and engagement initiatives across teams, as well as influencing how leaders think about retention in your company
  • Walk step-by-step through of lessons applied to dozens of examples from companies like Instagram, Zoom, Spotify, Everlane, Airbnb, Turbotax, Jira, Credit Karma, Blue Bottle
  • And more…

The Retention + Engagement Deep Dive is designed for growth professionals who are looking to zoom in on retention, either because their job is focused on retention, or because they already have an advanced working understanding of the quant and qual fundamentals of growth and are looking to build additional competency in retention and engagement.

Apply for the Retention + Engagement Deep Dive

About the Growth Models Deep Dive
The new Growth Models Deep Dive addresses an essential new skill and topic that every growth practitioner needs to understand. Your growth model is the essential tool that drives alignment, prioritization, strategic investments, metrics, and ultimately, growth. Without it, your team ends up setting faulty goals, focusing on sub-optimal initiatives, and running in opposite directions.

This program goes deep into growth models across companies. You will:

  • Learn how the fastest growing products actually grow (hint: the answer isn’t funnels)
  • Dissect how the fastest growing products like Uber, Slack, Dropbox, Stripe, Airtable, Instagram, Fortnite, Tinder, and others grow using growth loops
  • Learn the detailed components of 20+ growth loops
  • Systematically construct growth loops your product can use after analyzing the three qualitative properties of every growth loop
  • Assess gaps and uncover opportunities for growth by identifying, measuring, and analyzing your products existing growth loops
  • Complete a step-by-step walkthrough to build your quantitative model for a single loop and your entire product
  • Communicate actionable insights from your growth model to obtain buy-in from leadership and across teams
  • And more…

The Growth Models Deep Dive is designed for growth professionals looking to focus on growth modeling, either because their job requires modeling their company or product’s growth or because they’re in a leadership role. It’s especially useful for growth leaders looking to influence leadership, set a team’s direction, and rally colleagues using growth models.

Apply for the Growth Models Deep Dive

About the Growth Series
The Growth Series is a comprehensive overview of the key topics in growth. The program is designed to help you accelerate growth of your product, company and your career by creating a prioritized list of retention strategies, building your quantitative growth model, and much more. Plus, the Reforge team spent +100 hours collecting feedback, investigating new growth concepts with experts, and analyzing the latest strategies coming out of top companies to completely overhaul the content with new topics, frameworks, and relevant examples.

During the Growth Series, you’ll learn:

  • Going from understanding one or two pieces of your growth model to understanding how the entire system works together
  • Evaluating the key components of growth (acquisition, retention, monetization) and how they feed one another
  • How to construct a holistic growth model, bringing together all the components of the funnel
  • How to understand and evaluate the user motivations behind the levers in your growth model
  • Running a continual, self-reinforcing experimentation process to execute against your growth model and user psychology
  • Learn how to properly call, dissect, and analyze an experiment, plus implement the results across your team
  • And more…

The Growth Series is designed for practitioners who already know the basics of growth and are figuring out how to take the next step. Participants are assumed to have knowledge about A/B testing, ad buying, and other fundamental tactics, and are ready to take on the bigger challenge of thinking about the entire picture of growth and forming a coherent and compelling strategy.

Apply to the Growth Series

The Email Marketing and Notifications Evolution Inside Companies

“Stop sending emails like a marketer. Start sending email like a personal assistant.”

I’ve communicated this line in a lot of my presentations on growth, but I haven’t talked about in depth the evolution on how to get there. There is a clear evolution that companies follow in terms of evolving their emails and notifications, from not sending them at all, to sending one-off blasts to their entire audience, to creating a lifecycle, to having a holistic personalized messaging platform. Having worked on these systems at my last two companies, I thought it would be beneficial to outline these transitions for people earlier along in their process.

Phase 0: We Hate Email

“We hate email, so we don’t send it to our customers.”

Almost every company I have worked with has communicated this at some point in its early stages, and it’s always wrong. Except for very specific groups, people don’t dislike email; they dislike bad email. The path to figuring what email your customers would like to receive is largely to ask what kind of value do they see in my product, and can I deliver that value in email form.

Phase 1: Mass Promotional Email + Personalized Order Notifications
If you are a transactional service, you have to send personalized order notifications, so that is where most companies start. At some point later, companies start sending mass emails to their broader audience about certain things like new features, discounts, etc. This effort will show improvements in key metrics, but it is very unsophisticated.

Promotions train users to wait for promotions to order, decreasing profitability. Also, with this approach, marketing is assuming a cadence for the customer instead of adapting to the customer’s cadence (and ultimately improving customer cadence to increase lifetime value).

If you are engineering constrained, there are some simple optimizations to this approach that will improve your performance:

  • Develop additional emails intended to drive habit formation (instead of just timely purchase). Examples include trending items, item sales, new merchants added, recommended items
  • If emails are successful, test them as push notifications as well
  • Take existing confirmation emails, and add marketing messages to them (other things to buy, set up a re-order, most popular items, etc.)
  • Send every non-transactional email and tweak to confirmation emails as an experiment with an enabled and control group to prove impact on lifetime value vs. unsubscribes/push permissions/app deletions

Phase II: Moving to Lifecycle Messaging
In Phase II, these additional emails and notifications that have been successful in Phase I start to form an automated program that consistently drives additional engagement from customers. In order to address messaging fatigue, these email and notification templates are managed to a frequency per month based on the team’s expected value of a good customer. This frequency is not based on data, but if everyone used the product properly, what would ideal frequency look like. The goal is to use messaging to remind people of the service and reinforce the habit. They are paired with personalized discount emails intended to drive new use cases and increase frequency.

These emails and notification templates are also managed against each other, so messaging does not get stale. For example, if you have three templates outside of confirmation emails, and you sent template 1 last week, you would attempt to send templates 2 and 3 before sending template 1 again. Also, each week, these emails have new subject lines to present them from looking like the same email as the previous weeks.

Phase III: Holistic, Personalized Messaging
As the Phase II approach flattens in terms of the additional impact it can drive, companies shift toward a more holistic, personalized system. This is a considerable investment, which we made at Pinterest. Essentially, product and engineering determine for each customer:

  • The right content
  • The right time to send it (day of week and time or day)
  • The right amount (how many emails and pushes to send)
  • The right channel (email, push, or both)

This requires a team to develop a log of every email/push sent to a subscriber, when it was sent, when it was opened, when it was clicked, what downstream engagement occurred from clicks, and which template it was. All emails and notifications are run through the same experiment dashboard as product changes to understand the impact on all key metrics. From this, it needs to determine:

  • The best day(s) of week and time(s) to send messages to each user
  • A prioritization of the templates to send based on historical click through rates and/or purchase rates
  • How many messages per a generic time frame maximizes lifetime value of each user

This usually starts via a rules based approach, and eventually becomes powered by machine learning. If you lack enough historical data on a user to do this, for example new users, you group people who used to look like those users as a segment i.e. previous new users and look at the best performing approach for them. Email can no longer be considered marketing at this point. It is considered an extension of the core product.

The team also starts optimizing deliverability through choosing better message transfer agent partners and segmenting IP addresses for different templates to isolate issues. The team may also start investing in more advanced security measures like DMARC.

This is a considerable investment, which is why most companies only start building this once there are sending millions of emails a day with a lot of history from operating in the first two phases originally. At this point, companies know the value of email, and can justify the investment.

In my opinion, every company should end up at phase III at some point. The question is how long it takes to get there. This varies based on engineering constraints, scale, and how long it takes emails and notifications to flatten off in terms of additional engagement by the previous phases. Outsourcing this to a marketing technology company is also very problematic as it requires access to all of your user data, and any migration of data from system to system slows down performance. At a certain scale (like Pinterest), it is not even possible.

If you’re not at Pinterest’s level of sophistication, don’t dismay. Very few companies are. Just start to think about the long term evolution, and when is the right time to push for a step change in email and notification performance vs. continued optimization. It’s a big investment to shift from phase to phase, but the returns are usually worth it, and the impact of these emails and notifications in the current phase, and the struggle to improve their performance, should be what drives that decision to make additional investment to get to the next phase.

Currently listening to XTLP by μ-Ziq.

Four Strategies to Win Big with Low Frequency Marketplaces


Frequency creates habit which creates loyalty which creates profit. Uber and Lyft are successful because consumers need to get from A to B multiple times a day, forming habits that lead to long (and high!) lifetime values. Grubhub similarly benefited from people eating more than once per day.

But there aren’t that many business opportunities that have daily — or even weekly — frequencies. And those spaces have become very competitive. For example, how many food delivery companies can you name? Now add in groceries or meal kit cooking companies. All that just for the “eating” use case.

What if the natural frequency of use for a transactional business is low, like buying a house, selling a car, or booking a trip? How do you create a successful business if ideal frequency is quarterly or yearly or even once every few years? You would be unlikely to create a habit or loyalty, much less get the customer to remember your brand name. That is usually the case. If you don’t create loyalty, then you usually have re-acquire consumers when the need eventually arises again. This hurts customer acquisition costs and lifetime value. This fact makes building a successful business with low frequency extremely difficult.

With a low frequency business, you usually need to have a high average selling price to make up for the lack of frequency. While an order on Grubhub may cost you only $25, the average transaction size on Airbnb is hundreds of dollars. But a high average selling price alone is not enough to become a massively successful business. I’ve seen four distinct strategies for how to thrive in low frequency marketplaces. They all revolve around being top of mind when the transactional need occurs, no matter how infrequent that need is. I’ll start by talking about the most common approach, and then lead into some that are actually more valuable and defensible.

The Expedia Model (AKA SEO)
Companies that pursue this model: Thumbtack, Expedia, Apartments.com, WebMD

My first job was at Apartments.com. We were a classic low frequency marketplace. People search for apartments at most once a year, and there isn’t a whole lot of value you can provide in between apartment searches. So what did we do at Apartments.com? If you do not create habits or loyalty with initial use, users go back to the original way they solved the problem last time. Where do people go where they are searching for help renting an apartment? Usually, Google. So, the Apartments.com strategy was to rank well organically on Google so when people did search again for an apartment, they’d be likely to see us and use us for their search again.

SEO can be a very successful strategy, but the entire company has to be geared around success on Google. This strategy is also susceptible to platform shifts, like Google algorithm changes or Google deciding to compete with you. It also tends to shift companies toward portfolio models at scale. This is why Expedia owns Hotels.com, Orbitz, Hotwire, Travelocity, and Trivago, and why Priceline owns Booking.com and Kayak. When you rank #1 for your main keywords, the only way to grow is to own the #2 and #3 spots as well.

The Airbnb Model (AKA Better, Cheaper)
Companies that pursue this model: Airbnb, Rent The Runway, Poshmark

Sarah Tavel wrote a post about products that are 10x better and cheaper than their alternatives. You can definitely pursue this strategy even if you have low frequency. Airbnb was significantly cheaper than hotels, and many people, once they experienced Airbnb, found it a better experience as well. It was a more unique listing, in a “more real” part of the city, and they had a connection to a local. So, even though people only travel once or twice a year on average, when they do, they remember the Airbnb experience and start there directly instead of on Google, competing with the SEO behemoths of Expedia and Priceline.

Finding this level of differentiation in different industries is not easy, but worth contemplating. Airbnb is not the only startup that has entered a crowded space and grown rapidly by figuring out how to be 10x better and cheaper. RentTheRunway allows you to access high quality fashion without the high price, and without storing it, because dressing up is increasingly a low frequency occurrence.

The HotelTonight Model (AKA Insurance)
Companies that pursue this model: HotelTonight, One Medical, Lifelock, 1Password

There are certain businesses that are needed infrequently, but when they are needed, they are needed with great urgency. Example spaces include urgent care, being stuck in a random city unexpectedly, and fraud alerts. The key here is that someone keeps the app or account live despite a lack of usage because the fear of when it might be needed is so great. This is a hard strategy to pursue, but once the value prop is established, these companies remain sticky despite their lack of frequency.

The Houzz Model (AKA Engagement)
Companies that pursue this model: Houzz, Zillow, CreditKarma

Contrary to what many might think, keeping users engaged in a low frequency business is indeed possible: the key is a non-transactional experience. Many of these approaches have a “set and forget” component to them where they reach out with pertinent information in a more frequent way. Zillow is the first example I can remember that utilized this strategy. Even when not actively looking for houses to buy, Zillow kept users engaged by valuing their existing homes via the zestimates. CreditKarma reaches out with alerts and monthly credit check updates.

Houzz is a great example that is more recent. People remodel and redecorate their homes infrequently, but they are inspired more regularly. Houzz has a great product that shows home inspiration that can be saved and discussed, and when needed, but much more rarely, transacted.This is a product people engage directly with in instead of having to have content pushed to them

For this strategy to work, you essentially build a second product that enables frequent engagement — not a transactional product. Engagement strategies for low frequency marketplaces take advantage of an inherent human desire to stay up-to-date on things important to them. This won’t work for all industries. We actually tried this at Apartments.com, but were not successful because renters don’t care as much about investing in their living situation as homeowners.

A common confusion is that loyalty programs are an example of this. What loyalty programs usually do is increase frequency or target users that have high category frequency, like business travelers in the travel segment, rather than create loyalty from infrequent users. It is still a very valuable strategy, and I have blogged about loyalty programs if you want to learn more.

Of the four models I wrote about above, you will notice that not one of these is a brand model. Many of the sites listed in the SEO model have spent hundreds of millions of dollars building brands. Yet most travel searchers still start with Google. Brand is an extension of the Airbnb model, not its own strategy. If the product doesn’t deliver on a differentiated experience, brand building usually does not create loyalty.

So, if you’re building a low frequency business, do not dismay. There are many paths to still becoming a very large and differentiated business. These strategies are difficult but very rewarding if they are executed well.

Currently listening to Take Me Apart by Kelela.

Product-Market Fit Requires Arbitrage

One of the most discussed topics for startup is product-market fit. Popularized by Marc Andreessen, product-market fit is defined as:

Product/market fit means being in a good market with a product that can satisfy that market.

Various growth people have attempted to quantify if you have reached product-market fit. Sean Ellis uses a survey model. Brian Balfour uses a cohort model. I prefer Brian’s approach here, but it’s missing an element that’s crucial to growing a business that I want to talk about.

First, let’s talk about what’s key about Brian’s model, a flattened retention curve. This is crucial as it shows a segment of people finding long term value in a product. So, let’s look at what the retention curve shows us. It shows us the usage rates of the aggregation of users during a period of time, say, one month. If you need help building a retention curve, read this. A retention curve that is a candidate for product-market fit looks like this:

The y axis is the percent of users doing the core action of a product. The x axis in this case is months, but it can be any time unit that makes sense for the business. What makes this usage pattern a candidate for product-market fit is that the curve flattens, and does fairly quickly i.e. less than one year. What else do you need to know if you are at product-market fit? Well, how much revenue that curve represents per user, and can I acquire more people at a price less than that revenue.

If you are a revenue generating business, a cohort analysis can determine a lifetime value. If the core action is revenue generating, you can do one cohort for number of people who did at least one action, and another cohort for actions per user during the period, and another cohort for average transaction size for those who did the core action. All of this together signifies a lifetime value (active users x times active x revenue per transaction).

Now, an important decision for every startup is how do you define lifetime. I prefer to simplify this question instead to what is your intended payback period. What that means is how long you are willing to wait for an amount spent on a new user to get paid back to the business via that user’s transactions. Obviously, every founder would like that to be on first purchase if possible, but that rarely is possible. The best way to answer this question is to look with your data how far out you can reasonably predict what users who come in today will do with some accuracy. For startups, this typically is not very far into the future, maybe three months. I typically advise startups to start at three months and increase it to six months over time. Later stage startups typically move to one year. I rarely would advise a company to have a payback period longer than one year as you need to start factoring in the time value of money, and predicting that far into the future is very hard for all but the most stable businesses.

So, if you have your retention curve and your payback period, to truly know if you are at product-market fit, you have to ask: can I acquire more customers at a price where I hit my payback period? If you can, you are at product-market fit, which means it’s time to focus on growth and scaling. If you can’t, you are not, and need to focus on improving your product. You either need to make more money per transaction or increase the amount of times users transact.

Some of you might be asking: what if you don’t have a business model yet? The answer is simple then. Have a retention curve that flattens, and be able to grow customers organically at that same curve. If you can’t do that and need to spend money on advertising to grow, you are not at product-market fit.

Other might also ask: what if you are a marketplace where acquisition can take place on both sides? If you acquire users on both sides at the wrong payback period, you’ll spend more than you’ll ever make. Well, most marketplaces use one side that they pay for to attract another side organically. Another strategy is to treat the supply side as a sunk cost because there are a finite amount of them. The last strategy here is to set very conservative payback periods on both supply and demand sides so that in addition they nowhere near add up to something more than the aggregate lifetime value for the company.

Currently listening to From Joy by Kyle Hall.

What’s Your Mobile Loop?

One thing that is true about internet visitation habits and even more true about mobile visitation habits is the loop. The loop is the sequence of places you visit when you sit down at your computer or pick up your phone. What’s so interesting about the loop is how short it is for most people. With billions of web and millions of apps, most people’s loop only consist of a select few destinations. So, if you’re building an internet business, to be successful, you need to become part of quite a few users’ loops, or find a way to inject your content into the destinations that are in their loops. For example my mobile loop generally looks like this:

Gmail app (only if new content pushed there)
Feedly app
Twitter app
ESPN website
Amazon Kindle app (currently reading Sapiens)

My web loop looks like this:
Gmail
Feedly
Twitter
Quibb
Pinterest
ESPN
some niche sites for music

If you’re capable of becoming a part of people’s loop, you have what I call destination appeal. You are where people want to go to when they are bored. There is another way to have destination appeal, and that is to be very successful at branding, or to have a frequency not at the level of “I’m bored”, but still pretty high e.g. searching. I set the barrier there at monthly. For the former, Airbnb might not be in any of my loops, but it is the first place I go when I need a place to stay. For the latter, I don’t check Google unless I’m searching for something, but I search for things very frequently. See here for more musings on mobile apps and frequency.

If you are not in many people’s mobile loops, even if you can build destination appeal, you probably need to focus a lot of attention on injecting your content into apps that are commonly in people’s loops (Facebook, Instagram, Twitter, Pinterest). For some of these apps, this can be done organically, but much of it comes in the form of arbitrage i.e. paying less to be in front of users on that site than you will make from getting in front of them.

So, examine your loops. Is your company in your loop? If not, are you injecting your company’s content into your loop effectively?

Loyalty Marketing Part II: Making a Program and Keeping It Successful

Read part 1 of of my series on loyalty marketing.

In my previous post on loyalty marketing, I talked about the different types of loyalty programs, and how to identify which type of program your company should pursue. Once that happens, do you slap up generic version of a program that tackles your needs and call it a day? Absolutely not. Now that you’ve identified a program type to target, you need to determine a version that your users will respond, that will fit your brand, is profitable over the long term, and is future proof. Let’s tackle user response first.

Understand Reasons Why
Your can’t expect your users to change their behavior until you understand why their behavior is the way it is. Let’s say most of your users use your product regularly, but not every time they have the problem you solve. In order to create a successful program, you need to figure out why they don’t use you the times they don’t. The only way to do that is to talk to them. Take random people in the segment you’re trying to change and arrange a phone call. Reward them for it. In 20 minutes, with targeted questions, you can learn all you need to know about the time they’re not using you. A standard question to learn this is “Tell me about the last time you did X and didn’t use us.” Keep doing these calls until you start hearing the same types of responses over and over. In my experience, things settle around four or five reasons. For loyal, but infrequent users, it works very much the same. Talk to users, but this time ask “Why don’t you use X for [new use case]?”

Understand the People Behind the Reasons, and Pick a Reason
These phone calls don’t give you any statistical representation around how popular these reasons are for the broader audience that isn’t using you every time. So, now that you have your reasons, you can survey the broader group, asking them, “When you do X and don’t use Y, which reason best describes why?” and make the answer multiple choice with the responses you received over the phone. With a good enough response, you can now stack rank the reasons why people aren’t loyal to you. Some may be product changes you need to make. Some may not be helped. But, more than likely, you can address most of them with an incentive. You can go further down the phone calls + survey rabbit hole until you have full personas of users. Knowing the reasons why users aren’t loyal and what types of users you have can make you say, “I want to target this reason for this persona.” The same philosophy applies to incentivizing use cases. Our survey question is the same question you ask over the phone, except now it’s multiple choice. The goal again is to be able to say “I want to target this use case for this persona.”

Testing the Program
Now, you’re ready to build a program. At this point, it’s mostly a creative exercise leveraging psychology. Invent a bunch of a programs that might incentive these users, narrow down the ones that are most likely to incentivize users and be profitable, and test. Email is a great way to test different programs because you don’t have to build much and can book keep manually to get enough data without users knowing it’s not a real thing. It is also not a bad idea to run your users through these program ideas over the phone or in person, but remember that what they say and what they’ll do may be very different. Still, talking to them can prevent some gotchas.

Once you have a program, you need to test in a live way. Depending on what type of program you build, you may be constricted. For example, Yummy Rummy at GrubHub was considered a sweepstakes, so we could not legally have a control group. A control group is always the best way to test. Sweepstakes laws are at the state level in the U.S., so if you have two states that perform very similarly, that may work. If you don’t, you need to measure pre and post data. Pre and post data is not ideal for a few reasons. The main one is that loyalty programs typically take time to change behavior, and if you turn them off, it will take time for behavior to change in reaction to that as well. You don’t want to be running the program for a over a year, and not be sure if your pre data is still relevant. What typically happens in these scenarios is that programs are pulsed, like the McDonald’s Monopoly game being available for a limited time yearly. There is too much money being spent on a loyalty program typically to not know for sure if it is working or not.

Long Term Success
One other dirty little secret about loyalty programs is that they tend to ebb in effectiveness over time. Humans are motivated by variable rewards, and if your program is static, your users may become used to it, and it may not create long-term behavior change. That is why I recommend creating a variable program. At GrubHub, we made Yummy Rummy available every three orders instead of every order, and the reward could be anything from a free drink to free food for a year. Furthermore, if you lost, you got a consolation prize that was something random from the internet. But, I don’t think that is even enough. You should strive to think of your program as constantly evolving to stay interesting to your users. This will make your program stay effective for longer as well as give you the flexibility to tweak elements to make it more interesting to you as the business. I have seen many companies stuck with a program they no longer think is effective, but too afraid to shelve it because of potential user backlash.

The other advantage of creating a living, evolving program is that, if the original incarnation is effective, you can change it to move users further up your user lifecycle. For example, let’s say you’re trying to incentivize platform use in your original incarnation of your program. You might be very successful at that, and then find an opportunity to take those same users and get them to use the platform more by incentivizing use cases. Now, you can do that by evolving the same program instead of starting from scratch. Or, you might have taken loyal users and gotten them to use you for more use cases. Now, you can adapt that same program to build a moat around them. This all boils down to what a holistic loyalty program should look like in three steps for most internet businesses:
1) Build loyal users in one use case
2) Increase frequency by incentivizing use cases
3) Build moat around those users

This happens to be how most marketplaces or social networks grow into behemoths. They nail an initial use case, build a loyal user base for that, gradually expand use cases, and then work to keep those users locked into their platform.

Loyalty Marketing Part I: Strategies and Segments

There seems to be a lot of confusion about loyalty marketing and how loyalty programs work. To an outside consumer, I guess the confusion is understandable. Most loyalty programs are branded as a value to the customer, a reward for their dedication. Most loyalty programs’ primary goals are not to add more value to consumers (though when they’re done well they do that too); their goal is to create more value for the company. I’ll break down how to think about loyalty if you are a business that is wondering if a loyalty program makes sense for you.

The first thing to understand is that every business has a loyalty problem; it just might not be the loyalty problem they’re expecting. To make this clearer, I’ll split consumers into four areas. Depending on where most of your company’s consumers fit is where you’ll spend your effort in thinking of a loyalty programs. The first thing to do is split all of your consumers into loyal and non-loyal and frequent and non-frequent. Loyal is defined by doesn’t use a close competitor as well as you for what your product/service does. Frequency is a bit more nuanced. Your product/service should have a target frequency you’re setting. For Pinterest, that might be daily. For GrubHub, that might be once or twice a week. You then can build a 2×2 matrix like the one below.

Each of these four buckets requires a loyalty program targeting different actions by consumers. Just to be absolutely clear, let’s go through that exercise for each segment.

Frequent, Loyal
Action: Keep consumers doing what they’re currently doing

Frequent, Non-Loyal
Action: Get consumers to migrate usage of competitors to you

Infrequent, Loyal
Action: Get consumers to use product/service more

Infrequent, Non-Loyal
Action: Determine if product issue can increase frequency. If not, ignore.

Now, we should talk about these strategies in a bit more detail. I’ll skip infrequent, non-loyal since it’s a combination of two other strategies, and probably implies a product or market problem.

Infrequent, Loyal Strategy: Incentivize Use Cases
In this segment, consumers use your product loyally, but not enough to your liking. This implies that there are not enough use cases for your product in the eyes of the consumer. That could be because these use cases do not exist, or because the consumer doesn’t perceive them to be relevant. If the use cases do not exist in your product/service, you need to build them into your product. Take, for example, Homejoy expanding into all sorts of home services after starting with house cleaning. If the use cases do exist in your product/service, but consumers aren’t using them, you need to invest in awareness or incentivizing a trial of them. For Pinterest, this might be upselling someone who uses the service for recipes to try planning a vacation with the service, or a web Pinner to try the mobile app. For GrubHub, this might be giving a discount for a pizza orderer to order sushi, or a web orderer to try their first mobile order, or a delivery user to try their first pickup order.

These opportunities might not exist or be worth the effort. When I worked at Apartments.com, we knew people would only look for apartments once a year or less. There was not much we could do influence that. What we could do was stretch our product to be useful for not just the apartment search, but also services you need once you find an apartment e.g. moving. Beyond that, there wasn’t much opportunity we could tackle, meaning we’d probably have to spend money to acquire those same users whenever they looked for an apartment again.

Frequent, Non-Loyal Strategy: Incentivize Platform Use
In this segment, consumers are very active, but don’t always use your product/service over a competitor. This is the most common type of loyalty segment because it’s easy to understand the upside. You can typically measure how much activity occurs off your platform. Here, you need to invest in an incentive to move those uses onto your platform. This typically takes the form or rewards points or punch cards.

Frequent, Loyal Strategy: Build Moat
In this segment, consumer are very active and don’t use anyone else for your product/service. These are your best customers. So, a loyalty strategy shifts from trying to increase how often someone use the platform to doing all you can to make sure these consumers don’t decrease their use of your platform or are wooed away by a competitor. This is by definition a money losing strategy to decrease risk instead of a money making strategy in the first two segments. Moat strategies can take many different forms and are frequently misunderstood. Some start out looking like the same strategy as frequent, non-loyal. One common one is to increase switching costs. One example of that is Facebook shutting off friend access to competing apps.

Many other moat building strategies get much more creative. They rely on looking at every possible risk to your consumers doing less of what they’re doing today and trying to address it. One of the most ambitious is Google’s launch of Android. Google makes most of its money from web advertising. They saw consumer attention shifting from an open, web platform they increasingly controlled via their browser Chrome to closed platforms on mobile owned by competitors Microsoft and Apple. So, they acquired and put hundreds of millions of dollars behind their own, open operating system in Android, which they charge no money for, but powers most smartphones all over the world. This is all so they could continue to control how people searched and saw ads in a mobile world.

So, we can go back to that 2×2 with our strategies now.

Now that you understand the segments and their corresponding strategies, you need to identify where the opportunity is for your product or service. The easiest way to do that is to run a survey to determine loyalty, and mine your user data for frequency. Then, see where the highest percentage of your users are.

This post covered how to identify which segment to focus on and the appropriate strategy to pursue. My next post will talk about making that strategy and implementation successful.

Read part 2 of of my series on loyalty marketing.

No Replies Will Be Accepted to This Blog Post, or A Lesson on Customer Variability

A friend’s startup recently sent me a marketing email that was sent from a “noreply@” email address. I chastized him in an email about this, and he asked to justify why it’s a big deal. I responded with an email he probably was hoping was a lot shorter. I figured I’d adapt it to a blog post if it helps others.

No replies are not a good idea for big companies and especially startups because you want to leave open any opportunity for a user to provide you feedback or ask questions. Giving your users a no reply tells them you don’t want to hear from them and don’t think their opinions/issues are important. They are also a little more likely to end up in spam folders. CampaignMonitor has a great write-up on this, so I won’t belabor the point.

What I want to do is go deeper and talk about customer service and variability. There are five types of customer variability in services:
Arrival: when someone wants a service
Request: what someone wants from the service
Capability: someone’s ability to perform a task required of the service (particularly in self-service)
Effort: how much effort someone is willing to put in for a service
Subjective Preference: someone’s opinion about how a service should be delivered

In service, there are two strategies for dealing with customer variability: accommodation and reduction. Accommodation is accepting customer feedback/interaction wherever the user tries to initiate it (email replies, chat, phone, twitter, etc.) and adapting to it. This leads to the best experience for customers, but also costs more and is more complex to do as you have to monitor/staff multiple channels. Reduction is the opposite: defining one way customers can do things (prix-fixe menus, noreply, hours of operation, iPhone only). This is very cheap and simple to manage, but leads to a worse experience for customers.

A restaurant is a type of service that can either reduce or accommodate variability in many ways.

How they reduce variability:
Maintain and publish hours of operation
Manage inventory through reservations
Special prices at off-hours
Prix fixe menus to eliminate choices for customers

How they accommodate variability:
Self-service salad bars and buffets
Train workers to host, wait, tend bar, and bus tables depending on which type of work is needed
Allow guests to stay as long as they want
Allow customers to set tips based on service

Both are totally viable strategies, but for customer feedback, I find that reduction strategies just don’t work because you eliminate feedback from certain types of customers i.e. those that don’t fit your preferred method for communicating with you. Those are probably the customers you need to learn more about. Also, for a reduction strategy to be successful, it pretty much requires great training of your customers. This is really hard to do unless frequency of interaction is high or the penalty for breaking the rules is very severe. In an app or website’s case, it is easy to train someone how to use the app because hopefully they are a weekly active user, but much harder to train them how to deal with a problem with an order, as that (should) happen very infrequently.

So, not only should you think twice before setting a no reply, but also you should think about your strategy and whether you’re attempting to reduce or accommodate variability or your customers and why.

On Lagniappe

I grew up in New Orleans, and there’s a word in use there that isn’t known anywhere else. Mark Twain said it was “a word worth travelling to New Orleans to get”. It’s a French word based on a Spanish phrase, the type of co-mingling of language only possible in such a place as New Orleans. So, what does it mean, and why is it so powerful?

Everyone in New Orleans will define lagniappe the same way, as “a little something extra”. The real meaning’s a bit more complicated than that. For example, the other day I ordered from Leona’s on GrubHub. Every time you order, no matter what you order, they add in at least one of these mini-cupcakes they make for free. First off, they are delicious. Secondly, there’s no message about it, and you didn’t have to pay anything for it. It’s just there for your enjoyment. That’s lagniappe, that little something given to you for free that you weren’t expecting but is such a treat to find. This type of practice is common in New Orleans, but you rarely see it elsewhere outside of the extremely common baker’s dozen or the fortune cookie in Chinese restaurants, which, because they’re so common, are expected and therefore no longer really lagniappe.

This type of practice creates such an overwhelmingly positive impact on the experience, I don’t understand why the practice in business seems to stay limited to small shops in New Orleans and bakeries. A small gesture can go a long way in creating a memorable experience with your customer, and whether you deliver food, make doughnuts, or do management consulting, there is always a way to give lagniappe. So, what little something extra are you providing your customers? How can you give lagniappe?

Solving for the Clingy Girlfriend, or Don’t Say Goodbye to Your Users Post Order

When developing your website, the most important thing not to do, but the easiest thing to do is to forget you’re designing for users. And these users are unlikely to be all the same, unless you’re targeting a very specific niche. As you think about developing relationships with your different types of users, it’s important to figure out the different types of users you have and their motivations. It helps me to think of them like girlfriends.

For this post, I’d like to talk about a specific type of user, or a specific type of girlfriend. This is the user that’s been a few times, and they are into it. They love what you’re doing. They’ve professed their love to all their friends. They love it so much that they want to spend more time using your site. This is great, right? Exactly what you wanted. Check. Got this one covered. Let’s go after the users who are less engaged.

Not so fast. A specific type of problem can arrive with these types of users, if you’re lucky enough to get them (and you are lucky to get them). Let’s say your site is designed for “get in, get out” type of use, and it does that well. Well, these users have been there, done that many times. They want more. And you’re not designed for that. They’ve read the About Us page, liked your Facebook page, read your blog, all that. They’re actually thinking to themselves, “I like this site so much. How could I spend more time with it?” Sound familiar? You’ve got yourself a clingy girlfriend.

The problem with these types of users if you’ve designed your site so well for them that they outgrow it too quickly. There are some key ways to keep the engagement high without them over committing and burning out on you though. Coincidentally, many of these tactics are used during the onboarding process for websites/mobile apps, but they are just as, if not more effective, post-order for these types of users specifically. It is helpful to think of it as if you are continuing to onboard these users even though they are most sold on the product, and if there is no additional way to onboard them, you need to invent some. A few key tactics I’d like to describe for this:

“What you can do now” suggestions – Has the user completed an order? Don’t consider the session over. Tell them they can now that they’ve completed whatever the task was they completed. It’s very similar to a quality onboarding process. Some suggestions:

Invite friends
Fill out your profile
Visit your Facebook page
Go to your Twitter account
Read your blog

Great example (post-onboarding): LinkedIn
Status bar suggests other way to complete your profile.

Great example (onboarding): Dropbox
Checklists that are stricken through once completed.

“Post-usage” features – Another way to keep this type of user engaged is to create some sort of mini-experience post order for them or asking them to take a more active role in the site. They should be designed as ways to lose time. Some suggestions:

Games
Asking them to moderate content
Shuffle of non-personalized content
View what others have ordered
View statistics for the site for the day/week/month

Great example (post-onboarding): Quora
No picture. When Quora was inundated with new users, it manual review process for each answer was running way behind. So, for many users that had show good engagement, Quora began asking users to review answers on the sidebar of the home page.

Great example (onboarding): Hunch
Hunch asks random questions to get to know you better. They have thousands of them.

“Did you know/Ideas for next time” suggestions – This can be a great time to educate users on features that are for more advanced users, additional use cases for the site, or even company background. Some suggestions:

Download our mobile app
About us
Related content OR Expose different product category
Advanced feature tutorial
Charities you support

Great example (post-onboarding): Chegg
Chegg asks to plan a tree on your behalf after an order.

Great example (post-onboarding): Qwiki
Qwiki suggests different Qwikis to view after finishing a Qwiki.