Dynamite

Personalize product discovery for every shopper

Dynamite personalizes your store within 300 ms, so every shopper sees the products that match what they came for.

Illustrative Daybreak website for Espresso browser. Brew method: Espresso. The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21
  • Message for this visit
  • A matching example
  • Relevant content first
Espresso browser
  • Opened two different espresso coffees, Golden Hour and Night Shift
Illustrative example: Daybreak. Generated design previews.

A coffees page that opens on the right shelf

Daybreak’s coffees page has a “Brew method” menu: All coffees, Espresso, Filter, Decaf. The menu is a target with three approved settings, and it opens on All coffees until a shopper has viewed two different coffees of one kind.

The data behind every decision

Dynamite decides what each visitor sees from how they arrived, what they do on your site, and the data you connect.

Espresso browser

From every visit

  • Where they came from
  • Campaign and ad
  • Pages they viewedOpened two different espresso coffees, Golden Hour and Night Shift
  • Past visits
  • Device, location, time

Data you connect

  • De-anonymizer
  • Company and contact matches
  • Your CRM and product data
  • Custom connectors

Install on any website with one line of JavaScript

<script defer src="https://www.usedynamite.com/dynamite.js" data-site="your-site"></script>
  • Shopify
  • Webflow
  • Framer
  • WordPress
  • Squarespace
  • Wix
  • Next.js

Built for your agents

Dynamite can be driven entirely over MCP, so your agents create, manage, and optimize personalization themselves.

  • Claude
  • OpenAI
  • Any MCP client

Self-improving

Dynamite reviews performance across all variants and personas, so each iteration gets better.

You keep control

Dynamite doesn't generate new designs on the fly. You configure the possible designs on our dashboard, along with who they are for.

Illustrative Daybreak website for Espresso browser. Brew method: Espresso. The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21
  • Message for this visit
  • A matching example
  • Relevant content first

Illustrative example: Daybreak. Generated design preview.

One site. Different starting points.

A message that fits

Start with the need that brought this visitor here.

Espresso browser
Brew method: Espresso

Brew method: Espresso

Filter browser
Brew method: Filter

Brew method: Filter

Show the right example

Put the product, workflow or story they care about in view.

Espresso browser
The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21

The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21

Filter browser
The four filter roasts show: First Light, High Tide, Wildflower and Day Trip

The four filter roasts show: First Light, High Tide, Wildflower and Day Trip

A relevant next step

Choose the action that fits this visit. Measure product discovery against the original.

Espresso browser
All coffees, one menu choice away

All coffees, one menu choice away

Filter browser
All coffees, one menu choice away

All coffees, one menu choice away

The useful things first

Help each visitor find their starting point in the same site.

Espresso browser
The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21

The store shows its four espresso roasts, from Slow Morning at $18 to Sunday Drive at $21

Filter browser
The four filter roasts show: First Light, High Tide, Wildflower and Day Trip

The four filter roasts show: First Light, High Tide, Wildflower and Day Trip

  • Opened two different espresso coffees, Golden Hour and Night Shift
  • Opened two different filter coffees, First Light and Wildflower

Questions

Does this reorder the entire catalog?

No. It sets a filter you already have to an approved value, or changes which registered row leads a page.

How is interest inferred?

From the distinct products this browser opened on your site, plus the referrer or campaign it arrived with. A visitor who matches no version sees the original.

How are unavailable products excluded?

By your store. It renders the product cards, so its own sold-out handling applies, and Dynamite reads no inventory unless a data connector you host supplies it.

Who is Dynamite for?

Any website whose visitors arrive wanting different things: online stores, SaaS and B2B sites, and publishers.

Do I need to redesign anything or migrate?

No. It works with any website out of the box.

How long until I see results?

You should be implemented in a day and start seeing results within a week.

Do I get help setting up?

Yes. Every account starts with an onboarding call, and we help you set up your first personalizations.

Are there guides?

We take you through setup on your onboarding call and share what works for sites like yours. Dynamite also bootstraps the first setup itself.

How much management does it take?

Dynamite can bootstrap the implementation and drive itself. The more data you give it, the better it gets. If you want to hold control over exactly what is shown and when, you can. Otherwise, the agents are good at it.

Will it affect my page speed?

No. Dynamite runs at runtime, after your page has painted.

Where does the data live?

In a cookie in the visitor's browser, backed up and enriched on a server.

What can be changed?

Dynamite can be pointed at any component or control to replace sections, add new ones, reorder, filter, and more.

How fast is it?

Dynamite caches variants, so the decision engine can render new content within 300 ms.

Sell to every visitor like they are the only one.