Flowpeek in your store.
Works with any store. See the ready integrations, the data and how setup works.
WooCommerceOrders, events and visit context.See the integration
ShopifyConnect store data with Flowpeek analytics.See the integration
ShoperCheck what the store connection covers.See the integration
IdoSellSee the data and how the integration works.See the integration
PrestaShopFrom integration to reports.See the integrationLooking for marketing tool connections? See the full integrations catalog.
Set up in 48 hours. We configure everything for you.
Flowpredict scores every visitor's interest in buying on a scale of 0–100. Find people with a high score before they add a product to the cart, and pass them on to marketing automation.
30 days for PLN 0 · first payment after the trial · cancel anytime. We set it up for you within 48 hours.
Use case · before the cart
Browsing the product list
Comes back to the offer. Compares. Hasn't added anything to the cart yet.
The model scores the available behavior. The cart is one of the signals — not a condition for entering this scenario.
A tag for the identified contact
high_intent_before_cart
A segment for follow-up messaging in your marketing platform.
More remarketing options
Someone can browse your product list carefully, keep coming back to the store and still have an empty cart. Flowpeek lets you use a score of that interest to pick the contacts worth helping toward the next step.
It's an extra signal for marketing: behavior + model score + contact context. Your current tool still runs the messaging.
01 · Behavior
Product lists, further page views and return visits build a history of interest. You don't have to wait for an add-to-cart event.
02 · Flowpredict
A 0–100 score helps you pick out the stronger signals. Open the journey and see what's behind a specific person's score.
03 · Automations
Combine a score threshold with the scenario's conditions. An identified contact with the right consent gets a tag, a field or another supported action.
04 · Your platform
Use the segment in your messaging: help them choose a product, remind them of a collection or clear up doubts before they buy.
You stay in control of who's in the group and what the message is. Your own threshold, a purchase check and re-entry rules let you fit the scenario to your store.
See prediction in automationsThe Flowpredict score is a relative intent score, not a percentage chance of purchase. Flowpeek passes supported actions to the connected platform; you set up campaigns and any audience sync to an ad platform in your own tools.
Flowpredict in the app
See four different profiles from Flowpeek. Compare the score, activity and purchase stage to choose the right next step — from watching interest grow to changing your messaging after an order.
Cart without a purchase
Two visits, seven page views and an add to cart. A score of 71 helps you spot the interest, and the journey shows where this visitor stopped.
Connect the score with the stage of the visit
Check which products were viewed and set up a scenario for a high score without a purchase. This profile is anonymous — a contact action requires the visitor to be identified, the right consent and a connected integration.
Scores 40, 71 and 95 belong to different people. It's a relative intent score on a 0–100 scale, not a percentage chance of purchase. The model accuracy shown on the cards comes from a test on a balanced sample.
Intent doesn't stand still
A customer doesn't have to add a product to the cart right away. They may keep coming back to the offer and comparing products. Flowpredict helps you catch a change in intent at the next scoring — and you decide which score you want to act on.
Returning to the store, browsing the offer and further interactions add to the visitor's history.
At the next recalculation, Flowpredict takes the available set of behavioral features into account.
It can go up, go down or stay the same. It isn't a sum of points added for every click.
Your score threshold, the purchase stage and the contact conditions decide whether the configured action can run.
Example: with a threshold of 70, a score of 40 doesn't trigger the prediction condition yet. If the score later reaches the threshold, the rule can be met even without a cart.
Build a scenario with predictionFrom report to decision
The Top users report shows the people with the highest interest scores for whom the report doesn't include a purchase. Open their visit histories and see which offer they keep coming back to.
The offer someone views tells you what they're looking for. The Flowpredict score adds an assessment of their behavior, which you can combine with a category, a product or engagement.
The Flowpredict score can trigger an automation or be one of its conditions. Mark a recognized contact who has the right consent with a label, or tag, in your marketing tool. That way you can build an audience based on their interest.
Next to the score, you'll see a feature chart and the course of visits. You can check what behavior goes with a high score, how many people get a similar score and how the model performs in tests.
intent before the cart
A visitor may browse listings, compare offers and come back to the store before choosing a product. Flowpredict gives you a signal to analyze this stage. Combine it with the page being viewed to prepare follow-up communication for recognized contacts.
top users without a purchase
The Top users report surfaces visitors with the highest scores and no purchase included in the report. Choose a segment, see the scale of activity and open the journey. You get a concrete starting point for analyzing your offer and choosing a marketing use.
score and behavior
A number alone won't pick the right message. Set the score against the feature radar, visits and the offer being viewed. Repeated returns to one brand may lead to a different communication topic than a quick look through several categories.
prediction in your platform
In Automations, set a score threshold and additional conditions, then choose an action in the connected tool. A tag, a score field or a change in scoring helps you choose an audience for follow-up communication. The action requires a recognized contact with the appropriate current consent.
from the first scenario to assessment
Start with a situation that matters for your store: interest in a category or returning to an offer without buying. Check the Flowpredict scores, choose the conditions and follow how the automation runs. The scenario doesn't have to start with the cart. Later visits and purchases will help you assess how useful the selected group is.
Getting started
Go from setup to a specific question. During the trial, you'll see how this module helps in your day-to-day work.
01
Open the Top users report for visitors without a purchase included in the report. Choose a segment and go to specific journeys.
02
Check the pages and products viewed. Choose a situation where intent, together with the topic of interest, can help with follow-up communication.
03
Set the threshold, conditions and action for recognized contacts with current consents. Check the tag in your marketing platform and what happens next.
In your daily work
High intent before the cart
Combine the view of the category URL with an ML condition and prepare a tag for follow-up communication.
What you gain
You use the interest signal at the selection stage, before an abandoned cart ever appears.
Same topic, different priorities
Set the score against their journeys and split the score ranges in the automation.
What you gain
You prepare groups by topic and strength of interest, which you can assign different next steps to.
Contact scoring for marketing
The scenario adds a tag, saves the score in a field or changes the scoring in your marketing tool, depending on the available actions.
What you gain
The prediction becomes information used for segmentation and communication in the tool you already have.
Your plan for the first 30 days
Choose one group of interested visitors and see how to use prediction before the cart.
We set up Flowpeek within 48 hours. We also check data availability and model readiness for your store.
Review the scores for visitors without a purchase included in the report and the journeys of selected people. Find the offer that's generating interest.
In the consultation, choose the threshold, conditions and action for recognized contacts, or a way to analyze visits further.
Verify the tags in the connected platform. Decide how you'll assess the usefulness of the group and the effect of communication.
You know where you can see intent before a purchase, and you have a concrete use for prediction or a plan for preparing the data the model needs.
Questions and answers
No. It's a relative score from 0 to 100 that ranks recorded signals of interest. It is not a percentage probability of purchase.
No. The cart isn't a required condition for the scenarios described. You can combine a view of a product listing or another page with the available Flowpredict score and trigger an action for a recognized, eligible contact.
Visitors with the highest propensity scores and no purchase included in the report. In the table you'll see, among other things, the score, the device and the number of page views; the Journey button takes you to the selected person's history. An entry in the report doesn't yet mean you have their contact details for communication.
You decide which scores belong to each group in the report. You set the threshold that triggers an automation separately. Match it to your goal and your store's data: an example of 70/100 is neither a universal cutoff nor a guarantee of results.
Yes, through Automations. You specify which Flowpredict score should start the process or allow the next step. In the connected tool, you can add a tag, fill in a field or change the contact's score, depending on the integration. Before acting, the system checks that the contact is linked to the visit and has current consent.
The score helps you choose a group, but it doesn't dictate the offer. You can prepare communication that explains the product, reminds them of a category or helps them choose a variant. Match the action to the customer's situation and your store's economics.
These are two separate questions. Read the model's test result together with the description of the group tested. A test with a fixed ratio of buyers to non-buyers doesn't show effectiveness across all of your store's traffic. Assess how useful the score is based on later visits and purchases. Check the impact of communication using a comparison method agreed in advance.
We set up tracking within 48 hours. Prediction also needs suitable data and an available model; we check their readiness during launch. In the consultation during your first month, we help you choose a practical use.
From description to practice
Advanced · 8 min read
Combine the intent score with an identified contact and tagging. Plan the conditions, re-entry and how communication ends after a purchase.
Read the guideAdvanced · 13 min read
Read profiles scoring 40, 71, 95 and after purchase. Translate the relative score into conditions and tags, spot pre-cart intent and check a small test.
Read the guide30 days on your own data
Start a 30-day trial. We set up and configure Flowpeek in 48 hours. You get a ready-to-use dashboard and a consultation in the first month.
30 days for PLN 0 · first payment after the trial · cancel anytime
See pricing after the trialReports that complement each other
Combine the Flowpredict score with the offer being viewed and the visit history. Automatically tag recognized contacts who have the right consent in your marketing tool, even before they add a product to the cart.
A card for each recognized customer: visits, purchases and consents in one place. Check the context before you choose an audience or plan your communication.
Open a visit's history and follow the path from the first page to the cart or order. See where the visitor came from, which products they viewed and what they did, step by step.
