Restart your marketing after a crisis

The first step is a complete list of what was running before the crisis and what state each item is in now. Restarts go wrong when something nobody remembered keeps running, or when a paused campaign comes back on its old schedule without anyone reading it again.

Kite lists every recurring workflow on your team, including the paused ones. That covers scheduled social posts, weekly newsletters, recurring website updates, and anything that fires when another system sends it an event. It checks that each paused workflow has both its schedule and its event trigger turned off, because a workflow with only its schedule switched off can still run when an event arrives. It also gathers the posts, emails, and landing pages it drafted before the crisis that haven't been published.

On your website, Kite reads the pages a visitor sees first, such as the homepage, the pricing page, and any campaign landing pages. It flags banners, countdown timers, and headlines that would read badly right now. If you've connected LinkedIn Ads, Kite lists the campaigns and their status. For channels Kite isn't connected to, such as an ad account or an email tool it can't reach, it adds them to the list so you can check them yourself.

The result is one inventory with a status beside each item: paused, still running, or a draft that hasn't gone out.

It's safe to resume when three things are true: the cause is fixed and you've explained the fix in public, new conversation about the incident is clearly slowing down, and people who look you up can find your account of what happened. A date on the calendar can't tell you any of that, which is why Kite writes these conditions down at the start and checks them before each stage.

Kite checks the conditions with its built-in research:

• News and social conversation. Kite searches the web, X, and Hacker News for your company's name alongside the incident, and counts how many new posts and articles appear each day. A story that produced 60 posts on Tuesday and four on Friday is fading.
• Reviews. Kite reads new public reviews and separates the ones about the incident from the ones about your product. A sudden wave of one-star reviews that repeat a headline looks different from steady complaints that describe a real visit or purchase, and Kite reports the two separately.
• AI answers. Kite asks ChatGPT, Gemini, Perplexity, and Claude the questions a prospect might ask, such as "Is [your company] safe to use?" It records whether the answers mention the incident, whether they include your response, and which pages they cite. If the answers only quote the original news story, your own statement may need to be clearer or easier to find.
• Your own update. Kite checks that your public statement or update page is live and that it says what changed, not just that you're sorry.

Kite can also set up a standing watch on your brand name and new reviews, so you hear about a fresh wave of attention without anyone checking by hand.

Each paused message needs a fresh read, because copy that was fine before the crisis can sound careless after it. Kite reads each item in the inventory against what actually happened and sorts it into one of three groups.

• Send as written. Messages with nothing in them that touches the incident, such as a how-to article or a product tip.
• Rewrite. Messages that are still useful but carry the wrong tone. Jokes, urgency language like "don't miss out," and claims that now sound ironic all belong here. A payroll company coming back from a data breach shouldn't send an email that calls its product "worry-free," even if the email was written months earlier.
• Retire. Messages that can't come back at all, such as a giveaway announced the day the story broke or a campaign built around the feature that failed.

Kite rewrites the second group in your brand voice and shows you each original next to its new version. It also writes the first message of the restart. That message usually explains what went wrong, what you changed, and what customers can expect from you now, and it doesn't contain an offer. You approve the wording before anything goes out.

Turn channels back on in order of how much the audience chose to hear from you, and move to the next channel only after the last one goes well. People who signed up for your emails or who visit your website on purpose are more willing to hear you out than strangers who see your ad in their feed.

Kite usually proposes a sequence like this one:

• Your website and your public update. The update page is live, and the homepage and pricing page read correctly for the moment. Kite makes these changes on a site it hosts or through your connected content system. On a site Kite hosts, you can undo any single change later.
• A direct message to existing customers. A plain email from a person at your company that explains what happened and what changed. Kite drafts it and can send it from your team's Kite address or a connected email product once you approve it.
• Organic social posts. Useful, low-key posts come back first, and promotional posts come later. Kite drafts the posts with suggested timing and publishes them through your connected social accounts when you approve.
• Promotional email and offers. Your regular newsletter and sales emails return once customers have heard from you directly and responded calmly.
• Paid ads. Ads put your name in front of people who didn't choose to see it, and the comments under an ad can restart the conversation you just got through. They come back last.

The order changes with the crisis. If the incident was a pricing mistake that mainly upset existing customers, Kite may suggest holding the customer email until the refunds are finished and bringing back ads aimed at new audiences earlier. Kite explains why it chose the order it proposes, and you can change it.

When you approve a stage, Kite turns the matching workflows back on and publishes the approved drafts. Paused LinkedIn campaigns are turned back on by you, in your LinkedIn ad account.

After each stage, Kite reads the response and compares it with your results from the weeks before the crisis, not with the crisis weeks themselves. The crisis weeks are a poor baseline, because almost any number looks like a recovery next to them.

Kite looks at the measurements that fit each channel:

• Website traffic and signups from Kite's built-in analytics on a site it hosts, or from your connected analytics product.
• Email delivery and replies, including whether the replies are about the incident or about your product.
• New mentions and reviews, to see whether the restart brought the story back up.
• AI answers, to see whether the questions prospects ask now return your update alongside the original coverage.

If the stage lands well, Kite recommends the next one. If a post draws angry replies or a new article picks the story up again, Kite recommends holding where you are, and it can pause the workflows it turned back on and any running LinkedIn campaigns. Kite reports what it saw and what it recommends in the conversation where the restart began.

Once the restart is complete, Kite writes up the incident while it's still fresh. The record includes the timeline, the statements you made and where they appeared, what was paused and when, the order the channels came back in, and how each stage performed.

Kite also writes a short review of the restart. It notes which campaigns were paused too slowly, which messages needed rewriting, and which conditions turned out to matter most. That review goes into your team's wiki next to your brand voice and positioning. If something goes wrong again, Kite starts from that record instead of from a blank page.