Busy executives read thought leadership articles to get help with their most vexing problems. That’s why article titles often promise a payoff, such as “How banks can supercharge technology speed and productivity.” Or “Why Your Recommendations Fall Flat (and How to Fix Them).” But once an article makes such a promise, it had better deliver. If the first 80 percent recaps the status quo and the article finishes with weak and obvious recommendations, it will alienate readers and damage the reputations of the article’s authors and their firm. Unfortunately, that’s often what happens.

Here are some common recommendations that, as the physicist Wolfgang Pauli once observed, are so bad they are not even wrong:

(One test of a recommendation I sometimes use is whether the opposite might ever make sense. Reject change? Discard data-driven insights? Ignore your customers? Ensure top management opposition? None of these pass the test.)

However, authors fall back on no-brainer prescriptions for a reason. It’s hard to solve gnarly business problems for a single company; that’s why management consultants are paid big bucks. And it’s even harder to find solutions that apply to many companies but aren’t generic and trite. Every company’s situation is different, even those in the same sector, and one size rarely fits all. But we can square that circle by making sure that recommendations meet these three criteria:

New, better, or counterintuitive. That means just about anything other than what readers have heard many times before. We once worked with an executive leading the development of banking apps at a major IT vendor. In his article, he counterintuitively recommended not seeking top management support for a new initiative. Instead, he had found an existing client with the infrastructure and the data to partner with for a pilot, and only when he had a proof of concept did he take it to top management, which approved it. That saved him six months for the approval process and $60K for resources that were contributed by the client bank. This won’t be the right strategy for every project at every company every time, but it’s an interesting and different approach. Readers will remember it, and they will figure out when it makes sense for them.

Includes compelling examples. Always cite examples like that above to demonstrate that your advice works. Many first drafts come to us without any examples. So far as readers are concerned, those articles are just theory. A CFO told us that he’d once been impressed by a paper from a major consulting firm that advised a complete restructuring of the “Office of the CFO.” There was, however, no mention of how or where the authors had done this. So, he looked up their profiles. As he put it, they were clearly bright, but they’d been out of business school only a handful of years and had never worked in a finance function. “If I’d tried to do what they recommended,” he said, “I’d have been burned alive.”

Addresses barriers to adoption. Many recommended courses of action will elicit a reaction from readers that it can’t be done at their firm, or by them, or for a reasonable cost. Since the author is unlikely to discuss objections with the reader, he or she has to anticipate and deal with them in the article. For instance, generating data-driven insights about customers doesn’t necessarily need terabytes of third-party information and a team of data scientists. Some low-cost ways that companies can gather customer insights include online tools like Google Analytics and Google Trends, and customer feedback from online surveys and reviews. Starbucks, for instance, uses online surveys and feedback forms to gather customer input and continually adjusts its products, store experiences, and loyalty programs accordingly.

By focusing on recommendations that are novel and demonstrably effective, and by anticipating potential obstacles, you can provide the actionable guidance that busy executives crave while building trust and solidifying your reputation. Which, of course, translates to building your business. And that’s a good thing.