It’s trendy for marketers to BAM-KAPOW their visitors with pop-up pleas to “JOIN OUR MAILING LIST!” These animated boxes, powered by hot startups like SumoMe and OptinMonster, promise a firm ROI handshake. In exchange for a smidgen of JavaScript, you’ll receive a lifetime of customer correspondence. But hold onto that handshake a little longer and you might feel the ROI go limp with hidden costs.

Here are four considerations to keep in mind before employing an email prompt strategy of your own.

1. Prompts lose their effectiveness over time

Email prompts are a heroin hit; they start off strong and then subside. If you log 100 new emails per month, installing an email prompt will surge that count to 1,000. What a feeling! Of course, it won’t last.

Month two brings 750 subscribers, and month three 500. You panic and tweak colors, A/B test copy, and fiddle with the launch animation, but it’s a losing game. Eventually, your incoming emails stabilize at 300 per month.

Why are diminishing returns guaranteed? Because novelty drives response and, unless your website is growing rapidly with few repeat visitors, your prompt will stop becoming novel. But even under ideal situations, you can still expect diminishing returns. That’s because the ubiquity of email prompts makes them less novel every day, for everyone.

Diminishing returns by itself isn’t that big of a deal; it’s the risk brought on by diminishing returns that’s the problem. The risk, to circle back to the drug analogy, is that you become so addicted to these extra contacts that you leave your email prompt running forever, cluttering your UX and adding bloat despite questionable returns.

Tip 1: Draw a line in the sand before installing your email prompt. Decide in advance when or under what conditions you’ll remove the prompt.

Tip 2: Perhaps run email prompts only on special occasions, in conjunction with a contest or event. This keeps their novelty high.

2. Prompts impact your other marketing metrics

The point of an email list is to nurture visitor loyalty. So when you install an email prompt, you better make sure that your prompt isn’t undermining loyalty outside of your list.

A pop-up is still a pop-up, no matter how slickly devised. And by their nature, pop-ups will always rub some people the wrong way — measure that effect!

Don’t just monitor how many new email addresses your prompt produces each month. Measure how the pop-up is impacting your general audience metrics.

Email prompts can increase your bounce rate. A less tech-savvy visitor might get lost in the prompt’s UX, while a more tech-savvy visitor might switch to another tab.

Email prompts can increase your visitor churn rate. People don’t like being harassed with pop-ups, especially when they’ve already subscribed to your list or firmly decided against subscribing. Email prompts aren’t great at discriminating between subscribers and non-subscribers.

If you’re going to install an email prompt, keep a close eye on adjacent marketing metrics. Or test the prompt among a subset of visitors and watch how those visitors behave over time in relation to a control group.

Tip 3: Don’t just measure a count of new emails; measure how the prompt is impacting your other marketing metrics. Or better yet, run tests.

Tip 4: Target email prompts to visitor segments. If you notice that Twitter users don’t stick around, set your prompt to display to the Twitterati and no one else.

3. Prompts have extremely high opportunity costs

Here’s a thought experiment: An advertiser wants to run a pop-up prompt on your website with a big banner inside. How much would you charge them? A lot, right? Maybe you’d even turn them away. So why the double standard?

A pop-up prompt is far more valuable than a rectangle ad that’s shoved into the corner. If the going rate for a rectangle ad is a $3 CPM, then an ad inside a prompt has to be worth 10 times that sum.

Is advertising your newsletter to your own audience worth a $30 CPM? For some businesses, with a killer email funnel, maybe. For the rest us of, the email prompt’s real estate is more valuable than what it’s promoting.

Tip 5: Don’t consider your email prompt “free.” Assign a CPM to your prompt and calculate a cost per email. Is the prompt worth the price?

4. Prompts drive less valuable subscribers

A core tenant of digital marketing is that visitors who come from different places will, on average, behave differently. For example, your Google Search visitors might make larger purchases than your Facebook visitors.

The subscriber who joins your list from an email prompt will always be less valuable than the subscriber who enters through a traditional text box. Why?

For one, it’s more work to join through a static text box, so those entrants are more motivated. But outside of that, email prompt subscribers feel blackmailed!

Cole sees your prompt and assumes that he has to enter his email address or get kicked off your site.

Emma sees your prompt and thinks that if she subscribes, she’ll never have to see that pop-up again (poor Emma).

Remember the movie Glengarry Glen Ross? The cast spends all of their time complaining about bad sales leads. They want the good leads, the potential customers who are genuinely interested in their real estate. Don’t clog your newsletter with bad leads.

Tip 6: Monitor the quality of subscribers from your email prompt. How do their open rates and unsubscribe rates compare to non-prompt emails?

Conclusion: Tread carefully or stick to vanilla text boxes

The worst thing you can do with an email prompt is set it and forget it. Plan your prompt’s entrance and behavior with care and watch it carefully. The moment it stops pulling its weight, grab it by the shoulder and shove that prompt out of the game!

Adam Ghahramani is head of digital product for a creative agency in New York City. Find him at adamagb.com or make friends on Twitter (@adamagb).