Pitching To Your Audience

Feedback is one of the oldest forms of data, but today data controls almost everything that we do.

popularity

One of the most time-consuming parts of being a journalist is listening to enough people to get all sides of a story. Writing the story is often the easy part. What makes listening more difficult is that there are detail people and concept people, but few that sit in the middle.

Some people love to get down into the details about the latest feature they have just implemented in a tool, or what they think makes their product unique. Perhaps they just learned how a customer is using their tool and want to show it as a new use model. They spend a lot of time talking about something I am very unlikely to write about or will abstract into a sentence or two.

Other people are great at painting the big picture but when you press them for details, they just trust that their engineering team implemented their vision and it all works. It can lack credibility, but they provide great sound bites.

It can be difficult to reconcile these two extremes and give equal voice to such a broad range of people and views. But who do we write for, and what do they want to read?

Often, it depends upon the subject matter. There are times when what I write is short on details. In fact, the whole article could be summarized, as ‘well, it’s complicated,’ or ‘it depends.’ And sometimes these articles take off on social media, in part because it provides a framework for a discussion and potentially because it does not provide definitive answers.

Other times, an article will go deep and provide lots of guidance to designers or verification teams about the best way to do something, and it falls flat in terms of readership – perhaps because there are fewer people who can immediately benefit from it, or it seems mundane.

This ties into something I have been hearing a lot recently – whoever has the data has a distinct advantage. It doesn’t matter if we are talking about articles, or work-loads in a datacenter, or the latest advances in AI/ML algorithms – the earlier you can access the data, the quicker you can use that to your advantage, and then it becomes possible to build a competitive moat.

This is creating a big awakening across the entire industry. Consumer data that is being gathered by search engines, social media, or shopping sites was once seen as ensuring the ads they received were relevant and targeted. Now, a growing number of people are realizing that it is exposing more about them and their lifestyles than perhaps they feel comfortable with, and certainly do not want it shared with people they did not directly give it to.

While stores always have been able to look at what sells well and use that data to decide what store-branded products they want to introduce, it is seen to be more problematic when online retailers can see across a much wider range of products — especially products they never had to stock and promote. They have realized the value of that data and do not allow others to see it.

This is happening in our domain of semiconductor design, verification, and manufacturing, as well. Companies increasingly are realizing they need to keep more of their data close to the vest. In the past, it may have been details about a design, but today it is spreading to the point where many EDA companies talk about not having access to enough data to be able to make tools, or at least tools that target large audiences.

It will be very interesting to see how in-chip monitoring plays out. Who owns the data? Does the chip designer have a right to that data, or is the data owned by the person who integrates that chip into a board or a system? That data could enable the chip designer to improve their product, but may expose a lot of information about the environment in which it is being used. That data may then allow others to learn from them and to copy their success.

Who can you trust with your data and what safeguards need to be put in place?

Back to journalism – we can look back and see the numbers, but that says nothing about the usefulness of the information contained within the article. It is only by comments that are left, or personal messages, that we can gain deeper insights. I am not fishing for comments here, of course. That reminds me of the tactics used by another publication, where their writers were required to comment on other articles in order to make it look like they were popular. It just led to a lot of very fake comments, which in my opinion made them look silly.

But please do feel free to reach out if you feel so inclined. I am about to start working on a story that was suggested by a reader because they wondered why something is the way it is, and questioned if it could be done better. Feedback is important data that we use to provide a better product for you.



2 comments

Lee H Goldberg says:

Bravo Brian – It’s all to easy to forget one’s audience in the hyper-caffeinated environment of today’s tech publishing industry. I’ll share this with my colleagues. And your comments about on-chip surveillance are begging for another blog that explores this chilling topic in greater depth!

Joao Geada says:

Brian: Good call on the value of data and the increasing trend to creating ever deeper impenetrable silos. As you point out, there are many good reasons for this, but it is going to have to lead to a new way of working with EDA companies, as tools can’t be optimized or created in a data vacuum.

Leave a Reply


(Note: This name will be displayed publicly)