All Weekly Finds and Key lessons summarized in a comfortable audio version: https:\/\/soundcloud.com\/kevin-indig\/weekly-finds-week-38\/s-cEqRl4cbsuc I'm constantly tweaking this format, so if you feel like it's getting better or worse please let me know (comments are activated). If you feel like something is missing, also make yourself heard in the comments. Ahrefs is an interesting company in many ways. Tim Soulo, CMO of Ahrefs, explains some of the things they do that make them so successful. https:\/\/www.youtube.com\/watch?vl1g3W_u0zcg One part is the power of brand. Ahrefs doesn't use Google Analytics, they don't use Schema, they don't break their head about keyword density. Instead, they focus on building a valuable brand, which Google rewards. One of the biggest issues is still that companies create content nobody wants to link to. They write about topics nobody searches for and they're not making it interesting. Key lesson: Ahrefs also doesn't hunt users with popups. Instead of focusing on feeding their email list, they create content around problems and explain how to solve them with their product. Speaking of SaaS, Justin Jackson (founder of transistor.fm) wrote a great piece about pricing and whether you should charge more or less in general. The answer is, of course, it depends. Because that's always the answer. So let me tell you what it depends on: value. The idea of value-based pricing is the opposite of cost-based pricing. For the later, you take what it costs to produce something, add a juicy margin, et voil\u00e0! Pricing. Value-based pricing works with the belief that customers pay as much as they think a product is worth, not how much they think (or know) what it costs to produce it. I mean, the iPhone is a perfect example (costs a bit over $200 to produce, sold at $1,000). The problem is value-based pricing is that value is hard to quantify. Customer research, market research, and testing are the best ways to find the ideal price. Key lesson: Perceived value is hard to quantify and more a feeling. If your customers feel more productive when using your product, for example, it provides value. As Marketers, we should think more about what the value for customers actually is and how we can intensify that feeling. One example that comes to mind are Asana's animations when a task is completed (I wrote about their Land & Expand model recently). When we talk about testing and iteration, there is hardly a better company at the moment than TikTok. In a fascinating and binge-worthy article, Eugene Wei explains the history of TikTok. One of the key points is that the TikTok founders didn't need to understand American culture to create a product that's deeply engrained in it. Algorithms get so good that barriers like language, manners, or traditions don't need to be understood to be reflected. When Bytedance bought Musica.ly to turn it into TikTok, they glued the algorithm that powers Toutiao onto Musica.ly and gave it superpowers (think Thor getting his hammer). Key lesson: we officially arrived in the age of sticky algorithms. TikTok is not a social network. It doesn't rely on a people graph to create stickiness. Instead, the algorithm understands users so well that it knows exactly what they want (even more than Youtube). Google uses machine learning to rewrite meta-descriptions and Portent ran an analysis on 30,000 descriptions and found that 63% are rewritten by Google. That means only 37% of meta-descriptions actually appear in the search results as originally written. They also found that Google's rewritten descriptions tend to be longer than those of sites and sometimes even reach 320 characters. Key lessons: It seems Google is more likely to rewrite meta-descriptions on position 4-6 to give sites a boost and see if they rank higher\/get more clicks.Also, mobile meta-descriptions have a higher rewrite rate than desktop. Sam Underwood introduces a way to determine keyword seasonality with KeywordsEverywhere's new historical search volume index. The new feature allows you to export the search volume over the last 12 months (I personally tried it out, it's really good). Once established, you can group them by topic and use the information to tweak your internal linking, title tags, and content creation. You can, for example, link to pages or topic clusters that are currently in high seasonal demand. Or, you adjust your meta-titles to reflect the season and understand exactly when searchers are receptive to that. Key lesson: Link to pages that target keywords that are seasonally in high demand from the homepage.