About
Avinash is the Chief Strategy Officer at Human Made Machine and a Brand Strategy…
Articles by Avinash
Activity
301K followers
-
Avinash Kaushik shared thisMethodologies matter more than metrics. Challenge yours, please. Exhibit #4981: How VTC is calculated by Ad Platfforms. Every platform has slightly different variation of what the metric's definition is, with different (and configurable!) attribution windows, and dramatically different "viewability" requirements to compute the result. End results: A. It is very hard to understand performance in a single channel (and if that is solving for you or the Ad Platform). B. It is impossible to compare the performance across channels (chalk and cheese). C. Intelligently reallocate your budget. This leads to data-influence lowering, the undesirable "gut instinct" rising, and in the end the business losing. In this instance, I simply take the metric and the methodology out of the hands of the Ad Platforms, and do it myself inside my own company. You might choose a different path. Either way... torture the methodologies behind the metrics and God will smile on you. 💸 #analytics #marketing
-
Avinash Kaushik shared thisStopping to smell the roses today. 🐝 Riverside Park, Manhattan. 🙏
-
Avinash Kaushik shared thisWith statistics, is Google playing a game of "heads I win, tails you lose?" You spend thousands/millions of dollars on your YouTube brand campaigns and Google performance campaigns. You are super smart. You refuse vanity-adjacent metrics. You insist on a incremental impact signal from high sample Brand Lift Studies and Conversion Lift Studies - both Google will execute for you. 🎉 Then Google reports the conversion/brand lift labeling 70 -90% certainty a "good chance," saying 50-70% can be used "directionally," and even lower-certainty results can "still help future advertising investments." G Official guidance: https://zqi.me/gcertain, Who is this "good chance" good for? In their article, Google has one line saying advertisers should "interpret results to their own business needs, and risk tolerance." Wonderful, let's do that. Are Google recommended "good chance," "directional," also good for you? Intentionally or unintentionally, the answer is that it is good for Google. In my newsletter today, I've used a simple Bayesian replication model, inspired by biostatistician Steven Goodman, to compute three metrics that take the business owner lens at the results: In english... Chances that you will get even a tiny fragment (+0.01) lift if you spend money again, the replication risk of your venture, and the chances that the results will be at a confidence you can give them to your CFO. Google sells advertising. You own the risk. Help yourself, help yourself. I've also shared a working Excel model of all of these KPIs and a calculator you can punch in your latest Google Brand, Conversion, Search results to see what the Business Owner Lens says about your results. Don't outsource your tolerance for risk to the company selling you the ads. #analytics #marketing
-
Avinash Kaushik shared thisThere are many ways to get user feedback (AND actively listen). My favorite kind? Hacks customers build to get around company products/tactics/strategy. If someone's willing to go that far... Something's really off. There have been waves of complaints about the relevance/quality of Google's AI Overviews. The latest includes instructions for permanently changing your browser's use of Google (on the left), plus an entire one-page website to make Google simpler: https://udm14.com Such "hacks" feedback is frequently dismissed by companies as "oh, it's a silly little minority." In this instance, I suspect most Google users will stick to default cluttered experience. Still... There is a volume of feedback at which I see it as a canary protending something bigger off with the product strategy. Hence, in my practice I advocate for active listening to active hacking. I love using them to break through the natural closed off caves in which all senior leaders at all big companies end up. The outcome does not have to be to roll back. The hacks can inspire the next generation of innovation. Customers are not always right, but they do always have the right to complain... And, at some point I listen and recommend it is profitable for you to do so as well. #analtyics #marketing.
-
Avinash Kaushik shared thisIs your analytics strategy CRAP or CARE? One of my older slides, still deeply relevant, quoting Mikko. CARE seems so obvious. Yet organizational silos, whether cultural or built into the org design, push companies toward CRAP. So does treating Analytics primarily as data puking, sorry, reporting and dashboards. And so does the distance between "Analysts" and leadership. Companies often end up in CRAP unintentionally. In the end, this is not a data or analyst problem. It is a leadership, process, and incentives problem. Live in CARE. Obviously. #analytics
-
Avinash Kaushik shared thisAre you a "Reporting Squirrel" or an "Analysis Ninja"? Test yourself: What does this fascinating report get right? What might it get wrong? And what's actively poisonous? 🙃 We tend to use "reporting" and "analysis" interchangeably. That's a mistake. A simple way to frame the difference: Reporting asks: What information was requested? Analysis asks: What decision are we trying to improve? For the second one, there are a series of questions to answer and steps to take. Each helps produce a higher-quality decision. I used the table below as a jumping-off point to explore the difference between reporting and analysis, using a very important question: Where should you eat the best pizza in Rome? The approach applies to almost every area of your work. It can help you produce more relevant insights and deeper impact. So... what's up with the table below? 😉 #analytics
-
Avinash Kaushik shared thisConversion Rate is most often Orders/Sessions, which drives CONVERT NOW silly behavior. Change the definition to Order/Users, the incentive now is to understand the consumer journey. A small change in your Finance Dashboard, a big cascading change in cultural incentives. It is ok now for Users to have 50 sessions (or visits) or 3. In the end, you still get rewarded if they convert. Hence, you focus on identifying why segment one needs so many visits and why segment two needs 3. You'll dig into sources and motivations of the humans. The landing page "scent" for that specific intent. You might not push the promo into their faces, or you might. Then, and this is the best, you might leverage AI better - rather than pushing AI hard to treat every visit as a single night stand, you'll tell the AI "hey, understand the consumer journey, and at each step of it, deliver the relevant content to the user's intent (and only sometimes that content is an ad)". AI happier using full intelligence. Users are happier with increased relevance. CFO is happier with, what else, 😘 increased profits. All from a little shift in a metric's definition. Carpe diem. #analytics #marketing PS: If you follow my advice above, then the sweet, sweet attribution awesomeness that is in your analytics tool finally becomes magnificiently useful. To appease the LinkedIn algorithm, I'll do the non-user-friendly "link in comments."
-
Avinash Kaushik shared thisDoes your company measure up to this benchmark of true customer-centricity? Bravo Costco Wholesale! The text reads: "No regular unleaded gas available. Premium gas price has been reduced to the price of regular to accomodate. Regular will be available on 8/14. We apologize for the inconvenience." Honestly, few would have complained. Some would have gotten Premium, others might have gone to other gas stations. I'm confident not a single person would have canceled their Costco membership. Still. Costco went above and beyond the call of duty. The difference between saying you are "customer centric" or "put the customer at the center" yada yada and actually doing it. [PS: I noticed the other day that I've been a Costco member for 29 years - longer than my marriage, kids, and everything else!] Credit where credit's due: Costco Warehouse in Superior, Colorado. Image source, and thread: https://lnkd.in/gai7TcBX #strategy #culture
Avinash Kaushik replied to a comment
12h
Brian: That is certainly one way to get a better view. I agree. (While I'm using VTC as an egregious example here to make a methodology point, I deeply dislike VTC as an incentive for anything - Unless validated with trust but verify large scale experiment, and often it is not :)).
Avinash Kaushik replied to a comment
12h
Purva: Most platforms will give you the base data (directly, via api, via dv360 type bits, etc etc). Then were you get it, you apply your own standard definition. Where you can't get the data for whatever reason you A. if the metric is important to you, you cut the spend as there are always more fish in the ocean, or B. the you use an alternative better metric or methodology to judge.
Avinash Kaushik replied to a comment
14h
Neil: 😁 Strategy IS more important than everything else. Your advice is absolutley on point: What problem are we solving for? And, for whom?
Experience & Education
Publications
-
Web Analytics 2.0
Eyrolles
Si le Web, tout comme le marketing et la publicité sur Internet, a connu ces dernières années une véritable révolution, nous n’avons pas pour autant changé notre manière d’exploiter les données que l’on peut y recueillir. Avinash Kaushik, leader d’opinion dans le monde de la mesure d’audience Internet, présente dans cet ouvrage sa méthodologie innovante pour les Web Analytics.
Other authorsSee publication -
Web Analytics: An Hour A Day
Wiley
See publicationWeb Analytics: An Hour A Day is the first book by an in the trenches practitioner of web analytics. It provides a unique insiders perspective of the challenges and opportunities that Web Analytics presents to each person in your organization that touches the web.
Avinash Kaushik is a expert in Web Analytics and the author of the top rated blog Occam's Razor.
In this best selling book he goes beyond Web Analytics concepts and definitions to provide a step-by-step guide to…Web Analytics: An Hour A Day is the first book by an in the trenches practitioner of web analytics. It provides a unique insiders perspective of the challenges and opportunities that Web Analytics presents to each person in your organization that touches the web.
Avinash Kaushik is a expert in Web Analytics and the author of the top rated blog Occam's Razor.
In this best selling book he goes beyond Web Analytics concepts and definitions to provide a step-by-step guide to implementing a successful web analytics strategy.
Recommendations received
35 people have recommended Avinash
Join now to viewOther similar profiles
Explore more posts
Explore top content on LinkedIn
Find curated posts and insights for relevant topics all in one place.
View top content