AMA: Quizlet Director of Product Marketing, Caroline Walthall on Messaging

I don’t have a magic bullet for this, because it can be a real challenge to align all your stakeholders ahead of a launch. At the end of the day, there are few habits you can learn to make things run smoothly: 1. Make sure you have a clear DACI framework with an executive sponsor to back you up. Ideally, product marketing is empowered to drive and own messaging. Oftentimes your head of marketing will be the approver, but keep in mind every organization will be different. 2. Decide who is consulted and informed in the approval chain. Oftentimes PMs, designers, and other stakeholders will be consulted, but those folks should not have veto power on the messaging unless something about it is inaccurate or far off from the brand message. Your CMO or head of marketing will likely be curious about it and may want to give detailed feedback. At some organizations, the CEO will want just an inform, at others they'll want approval sign off. 3. Create clear expectations for the order of review and sign off and set "review by" deadlines. I like to put mini checklists at the top of messaging docs and let stakeholders check their name off when they have signed off ahead of their due date. Once the doc is good to go, I write a note at the top of the page saying that the messaging is in a “locked” state, and I add a timestamp. I disable comments at this time. I’m sure there are numerous other ways to use more sophisticated tools for this process, but you can typically steer the ship quite well with a focus on role clarity, clear comms around the sequencing of approvals, checklists, and deadlines.

I'll give a spicy answer here. Build your own messaging framework. Sure you can borrow from things you see online, but you'll learn a lot about what feels essential to include when you work through the process on your own. Frameworks only get you so far. You have to know what you need for your particular launch, your company, and your stakeholders. I rarely use the exact same messaging framework, but I usually include a few consistent things: 1. One headline message summary. 2. Three value prop pillars. Doesn't have to be three, sometimes it's four, but the rule of three works well here. 3. Products, features, and reasons-to-believe that ladder to those value props. 4. Starter headline copy and several alternate sub-headers focusing on different positioning angles or parts of the product offering.

It depends how many cycles you can afford to spend, but my ideal flow is a five-step process: 1. Draft positioning strategy and possible positioning angles 2. Draft creative copy with the support of AI tools, colleagues, and copywriters 3. Test different succinct messages against each other in a market survey, ideally using a methodology like MaxDiff 4. Revisit your Strategy and Messaging Framework, with the full marketing funnel in mind 5. Test messages in product or in ads I covered most of these steps in other answers, but the part that I haven’t mentioned yet is how to revisit your strategy and messaging framework with the marketing funnel in mind (bolded above). What I mean by that is, some value props and messages will be a stronger fit for a consideration stage, vs. a trial stage, vs. a purchase conversion stage. If you spend a good amount of time conversing with your customers and prospective customers, you’ll likely have some intuition on this. We are currently refreshing our messaging and I’ve just gone through this sorting exercise. Here's how it went: * Top of Funnel: As I was thinking about our brand level messaging, I reflected on what I know about some of the common “objections” or reasons for not choosing to use Quizlet (at least in our priority markets). In deciding on which of the top messages would be strongest in building awareness and consideration, I wanted to choose value props and benefits that combatted some of those perceptions among non-users and lapsed users that Quizlet is not good for certain subjects. This helped me pick brand level messaging that was more expansive and told a story that goes beyond our typical subjects served and beyond our current target customer. * Middle of Funnel: Then, once you’ve garnered initial interest with a bigger picture narrative that expands what you're known for, you build your mid-funnel messaging to drive engagement. For this, I leaned more on past A/B test results in our product to hone in on clear, simple, functional messages and CTAs. We've found that it's important not to oversell at this stage and to mainly do whatever you can to get out of the customer's way. * Bottom of Funnel: Lastly, no messaging journey is complete without an actual purchase conversion. Since Quizlet is a freemium business, we need to focus on distinctive premium messaging that sells features and benefits that you can’t get with the free product. This type of messaging often has a stronger promise built into it, with clear and compelling reasons-to-believe. Once you’ve scaffolded your full-funnel story, you can still continue to do testing to validate your hypotheses, but it’s also important to make sure to look at it all together to make sure that it coheres. There are absolutely scenarios where your messaging should be pretty similar through all those phases. It depends a bit on your audience, your product purchase flows, and your business model. 

When messaging in a known category, there are two ends of the spectrum: differentiate or blend in. The best approach is usually finding some blend of the two. As marketers, we’re often taught to focus on our “UVP” (unique value proposition) or “USP” (unique selling proposition), but there are absolutely times when the desire to show up as a “special snowflake” can end up making it hard for consumers to grasp what it is you’re offering. To balance this, you may consider using naming and messaging that’s quite similar to your competition, especially if those products are looked upon favorably and have high adoption. To illustrate how to find the right balance of known and unique messaging, I’d like to share an example that is admittedly more focused on product naming, but the positioning and messaging is inherent in it. * When Quizlet first launched its textbook solutions offering, our biggest competitor in the space was Chegg. Chegg was well-known for having textbook solutions and it was a strong subscription driver for them. * On the flip side, especially during the pandemic the media was putting a spotlight on companies like Chegg that appeared to profit from students short-cutting their studies. Quizlet has always put academic integrity and actual student learning at the forefront of its product design and we were concerned about being associated with a brand that was in a tacit way, encouraging the opposite. * So in 2020 when we launched our version of “textbook solutions” (acquired from Slader.com), we called the offering “Explanations” because we wanted to emphasize the step-by-step nature of the content form and put the emphasis on understanding. * While we saw some early success and eventually built up credibility, it wasn’t until we changed the name to “Expert solutions,” that we saw a large shift in adoption and recognition among students. * My interpretation of this is that using familiar language goes a really long way, especially for top of funnel consideration. Chegg and several smaller competitors had called this product category “textbook solutions.” When we re-installed the familiar “solutions” noun, and added “expert” be our differentiator (both compared to Chegg, and to our own user generated content), we found the right balance of familiarity and distinctiveness that drove both trust and interest.

I think it's important to consider your both your conviction level and your interest and need to defend your messaging choices. Sometimes it's worth shipping changes without quantifying the exact impact because you know it's the right thing to do. Most of the time, however, if you're intellectually honest with yourself, you aren't so certain. Testing can be huge for clarifying the most important insights which helps you build a solid foundation for your messaging that is data-backed and defensible. I'll elaborate a little further below: * There are absolutely times when you as a PMM know your customers so well that you can confidently ship messaging updates without building out all the infrastructure to measure impact. At Quizlet we call these “just do it” projects. If the conviction is super high and/or the risk to impacts to metrics is relatively low, you can save yourself a few cycles and just make the change. * That said, if you believe strongly in the message change and you think it brings you closer to your customer and helps you better connect their problems to your value offered, it behooves you to find some way to account for that. This is where quantitative A/B testing or other forms of holdout or comparison groups can be helpful. If you can quantify the lift in conversion your messaging delivers, you build your credibility and help educate your organization on the power of strong consumer messaging. This is great for building your career and helping stakeholders see how copy alone can be a strong conversion driver. * If you don’t feel certain in what the best message might be, I usually ask my team to build out a collective Positioning/Messaging Hypothesis Doc. We’ve gotten in the habit of refreshing this a few times a year. Once we’re all looking at some of the same insights, value props, and style notes, it’s easier for us to start to sequence a series of A/B tests that help us learn what language, style, and CTAs resonate best. This type of testing can be a little laborious, but it ends up building your conviction as you layer and learn from past test results. This ultimately helps you feel comfortable with a less-is-more consistent message, without fear that you’re missing something or focusing on the wrong benefit. Beyond that, anytime you have data to back up your POV, you are in a stronger position to defend your choices with PMs, designers, execs, and other stakeholders.

At the end of the day, A/B testing messaging in the wild will get you the furthest, fastest. There are several ways to do this. 1. You can argue for some test budget to test messages in paid ads. Especially if you’re planning on using this messaging for brand marketing or demand gen, it makes sense to validate it via the channels you plan to use. 2. You can also A/B test messaging on your site itself if you get high enough traffic to run your tests relatively quickly. Most companies have A/B testing tools available, but if you don't I highly recommend advocating for your eng team to help build a simple self-service codeless copy test tool to enable you to run tests without significant product and engineering lift. 3. If you don’t have the luxury of being able to test messaging in situ, the next best thing for getting quantitative insights is validating messages through a market survey to your target audience. If you use a statistical technique like MaxDiff, you can get stronger forced rankings of the best and worst messages than through other survey formats. You often need an advanced survey tool for this, but it lets you increase your confidence on the most resonant messages and value props within just a week or two.

* A few rapid fire thoughts on writing clear and compelling messaging: * Start by writing your messages in simpler plain English and don’t worry too much about how it sounds yet * Ask yourself, “so what,” and “who cares?” Does it pass the sniff test? * If not, make sure the message taps into at least one value prop or benefit that matters to your target customer * If not, make sure the message ties to an insight you know about your users * If not, ask yourself if that feature or product or positioning angle is even worth mentioning, 50-75% of the time it won’t be * Edit and cut ruthlessly as you go * Put your simple messages into an LLM or send them to a copywriter for alternatives * Start to become more obsessive about word choice and removing extraneous words * Squint and make sure you’re not getting too “cute” or “technical” or “insert adjective that takes you further from the way real people think and talk” * Get feedback on it from prospective customers

It’s likely no secret that generative AI can be a helpful thought partner and writing sidekick to any PMM. If you haven’t started experimenting with ways to integrate it, there is no time like the present. In the last few months, I’ve been loving using the “Copywriter GPT” on ChatGPT to ask for lists of alternatives to basic messages that I like the idea of, but not the sound of. What’s so great about this, is that you can reinvest the time savings you gain in copywriting and copyediting further upfront in the strategy process. This frees you up to focus on the uniquely human decisions that go into successful messaging, which include: 1. What features and products are we going to feature as our cast of leading characters 2. What benefits and value props do we want to tell a story around that those features fulfill, and 3. How do we want to uniquely position ourselves in the landscape of alternatives? I find that spending enough time on those questions can get you to a pretty decent starter framework with a series of hypotheses and positioning angles to consider. * Once you’ve got some of those angles listed out, you can start drafting short form copy and asking your favorite LLM for alternatives. This usually helps me work my way towards something that feels more complete, or at least punchier. * You can try re-prompting ChatGPT or your chosen LLM with further style notes from your brand voice and tone guidelines or just with insights about your target customer. I am guilty of asking ChatGPT to write in a way that would appeal to Gen Z, and though I asked it not to lean too much into slang, some really hilarious stuff came back. 😂 Make sure to use your judgement! * Ninety percent of the time, I still apply some editing to what has been given back, but AI can help work me towards synonyms or turns of phrase I might have overlooked. * Lastly, once I’ve assembled all my alternative messages, I like to run some form of quantitative or semi-quantitative message testing to hone in on the best positioning angle(s).

* A few years ago, we had a bit of long-running internal debate about whether our target audience was more interested in “customization” or “guided study.” There was a product philosophy that we should offer guided options as much as possible to reduce friction. While that premise made sense, we often heard that the students who actually converted to our subscription offering, Quizlet Plus, were motivated by the promise of being able to create, customize, and modify their content, and also the way they studied. * Around the time, I conducted some mixed qual and quant testing that included a number of qualification questions to segment by age and persona. When you conduct this kind of unmoderated study with a wide enough sample, you can start to see a wide spread in how different individuals are in their interpretation of your message. * We saw that a lot of students were interested in the idea of “smart guidance to make the most of your time,” but also many of those same students had skepticism about the promise involved. The “guidance” piece sounded a little heavy-handed to many of the less-serious students that we had thought it might appeal to. We heard things like, “I don’t need someone to ‘guide me,’ I trust my own study skills. It could be helpful but it’s the least interesting option.” * Another variant we tested in the same study was “customize study sets”. This variant was expected to go over well with the serious students who were currently most likely to subscribe, and it generally did. But less serious students also resonated with it. They read it less as “something that would be hard to do,” and more as, “a way to target only what I need for a specific test.” * This insight enabled us to shift away from the initial ask from leadership to make “guided study” our topline value prop for Quizlet Plus. The data gave us conviction around centering the main message around “customizing sets to be exactly what you need.” The insight here was that customization isn’t always about going above and beyond as a power user, sometimes it’s just about relevance and empowerment, which is something that the “guidance” message failed to communicate.