Zuckerberg wears $800 tshirts amongst other carefully selected pieces of his wardrobe. He isn’t rummaging through a local Target for a basic tee and jeans. I think this is a good example of countersignal interpretation!
472 karma · joined October 31, 2011
https://reebz.com
>The work was often very low value-add to the business (often compensating for incompetence up the management chain).
The chain is what is important. This is not a data science specific issue.
Swap in or rearrange any team names to the original list and the last position or two will find this article true.
The fixation of hatred on marketing is one aspect: promotions. Of which advertising is a subcategory.
Another aspect, that most do not hate, is product.
Marketing needs to be reframed by the technology community as market-making between suppliers and customers; and not be as reductive.
If you’re a product manager, you’re fundamentally doing a marketing function. Only in recent decades have product managers been placed within engineering orgs to facilitate communications, which in turn has spurred some to perform revisionist history.
The reality is that budgeting and roadmapping is roughly an annual process. At that time scale and with sufficient people scale, management will need to illustrate how they will be spending millions of dollars and the sequence of activities.
Gantt charts are often the defacto communication method because people from many skills backgrounds easily understand them.
The service is free and only takes a 5% of subscription fees when they are active.
Most complaints are about intrusive advertising. Product management, customer analysis, demand and financial modeling, and all similar aspects of market making is what marketing traditionally is.
Advertising and promotions is a small component that tars the broader industry.
The great misunderstanding of this industry is that what ads you see are just as important as the ads you don’t see, which allows for profit maximization.
In your circumstance, it may be true that bank transactions are farmed and analyzed to be additive to data gathering on the web, which in turn is additive to the offline data gathering offered by various companies. Every transaction made is analyzed for location, merchant, purchase category, amount, date & time, and more.
Most of the other commenters who are claiming and linking to articles that ads don’t work aren’t in the industry. This is an important distinction because what we’re talking about here are the machines that crunch all this data are considered trade secrets and/or intellectual property. Also, slightly off topic, but it’s important to understand that ads != marketing. Ads are a child component of marketing. When you look at the overall function of marketing (of which product management is a discipline, no matter how desperately some attempt to align it to technology), then consider how much of a factor data analysis and operationalization of data is a driver of success for Fortune 500 companies. Almost any company in consumer technology (e.g. FAANG), consumer healthcare (CVS), telecommunications (AT&T), FMCG or retail (Walmart, Costco), and retail banking (Chase, Wells Fargo) are using these techniques to build better products, sharpen communications, and win new customers.
Anything must be done with authenticity. The book is really clear and even addresses that you simply can’t do it all the time. It illustrates the need to take the time to build relationships and, critically, listen. Don’t fake it.
Avoid any “SparkNotes” version of this book. It deserves a full read/listen. Although it’s decades old it is pragmatic and touches on topics like social proof and network effects before they were mainstream.
Context: my career went dev (not long) to consulting to product (most time) to executive management (current).
We're not talking about "Not Hotdog" here. Application of a model in a real-world at-scale scenario is a lot more than running inference and walking away.
At a bank credit decisions are evaluated by humans, frequently and often. These reviews are conducted in the forms of sampling audits, control processes, and other scenarios that involve internal bank employees and external regulators. In each case, humans will inspect the details of what occurred. This would be impossible with any type black-box model (SSL, deep NN, etc.).
Fine tuning labels black-box style is a terrifying concept to most who are working in fields where great risk must be managed to avoid unintended and disparate impact. Facebook making an oopsies suggesting my friends face in a photo instead of mine with a SSL trained model may seem trivial, but this non-human oversight is concerning for other applications.
If you were rejected by a bank for a home loan, you would want to know why your creditworthiness wasn’t evaluated positively (and banks must explain precisely why to regulators). And if a self-driving car made a decision in a collision, or a CV model performing cancer screening in X-rays that gave a bad result - ...or recommending politically divisive content... - or a thousand other real world examples that have human impact.
I have followed people who I find interesting and groups that align to my interests - but the loudest figurative voices are still the ones who pop my notifications most often. Frustrating, but I’m going to keep at it through February at least.
I’ve also had to rebrand it in my mind, it’s not social media. It’s more like community talkback radio and has found a niche in my listening habits where I’ve given up a few hours a week in podcasts to listen in on some wacky chats.
1. Much of the context is abstracted away when charts and summaries are crammed into a (usually) single page. This makes it hard for people to interpret quickly and requires a high degree of data literacy
2. Data is presented that fits retrospective summaries, or what I’ve learned to know I need to look at. This usually means the dashboard is not suitable for answering new questions, putting strain on the data model or falling back to custom queries
3. Combining 1 and 2 means that the people who build the dashboards are the ones who end up using them the most. So I find them good for ops teams and similar, but fall short of the coming promise of self-service data exploration yadda yadda.
...Which means the VP still needs a monthly hand-holding meeting to review these “self-service” dashboards (or worse - you screenshot and put it in a PowerPoint)