About

Managed behavioral personalization, for teams too lean to operate another tool.

Most personalization software is sold to people who then have to run it. Fulcrum is the opposite arrangement: we configure it, we operate it, and we prove what it did.

Start Your Free Pilot

The arrangement

Three things we decided differently.

Each of these is a structural choice, not a feature. Together they are the whole argument for handing personalization to someone else.

We do the work

No dashboard to learn, no campaigns to build, no rules to maintain. The managed service is included at every level, because a tool nobody has time to operate is a tool that becomes shelfware.

We cover the anonymous majority

Recognition begins at the first pageview via a first-party cookie, so personalization reaches the 90-95% of visitors who never identify themselves, not just the logged-in few.

We prove it against a control group

A permanent control group is what every number we report comes from, so the lift you see is a difference against untreated traffic rather than a total we take credit for.

Coverage

Every algorithm, every tier, from day one.

Most of the comparable set sells recommendation capability in modules, so a client pays more to add algorithm types. We run the other direction. Price scales with traffic volume, never with how much of the product you are allowed to use.

8: Algorithms, live from day one Every customer, every tier
0: Algorithm types gated behind a higher tier Not modular
24h: From new product to live recommendation Nightly recompute
0: Dashboards for your team to learn Fully managed

The suite is the same whether you are on the smallest page-view tier or the largest. That is what makes the pricing and the guarantee coherent: you are buying an outcome at a volume, not a permission set. See the eight algorithms .

Freshness

Rebuilt nightly, not refreshed on a campaign cycle.

Every recommendation relationship across your catalog and purchase graph is recomputed from scratch each night. Nobody re-runs anything, and nothing waits for a quarterly optimization pass.

A new product surfaces within 24 hours

Add it to the catalog and it can appear in recommendations by the next morning, with no manual mapping and no campaign to rebuild.

An out-of-stock item drops out

It leaves recommendations by the next morning, so you are not paying attention to a slot that sends shoppers to something they cannot buy.

A price change moves what counts as similar

Similarity shifts automatically with your own pricing, because the relationships are derived from current data rather than a snapshot taken at onboarding.

Measurement

We would rather show you the mechanism than the adjective.

Holdout lift is the number we report and the number the guarantee pays against: measured against a randomized holdout, taken conservatively from the lower bound.

We wait for volume before we evaluate

A minimum threshold of orders and views has to clear before any evaluation begins, so the earliest and noisiest data is never what triggers a call.

One evaluation, at term end

Calculated once, at the end of the term, at a 95% confidence level and reported from the conservative lower bound; there are no interim checkpoints.

Read the full method on how we measure impact , or what happens if the number falls short on the guarantee .

The company

Who you are dealing with.

Fulcrum SaaS is a personalization, recommendation and attribution platform for ecommerce brands, founded and led by John Golinvaux and based in Denver, Colorado. We sell one thing, Personalization as a Service, to Shopify Plus and BigCommerce merchants, and to the agencies and consultants who place it with their clients .