FAQ2021-02-22T13:57:28+01:00
What about GDPR?2020-10-27T09:44:39+01:00

We are 100% GDPR compliant. Not storing any user data and all data is Pseudonymization and hashed in our system only allowing clients to recognize their own data.

What distinct feature set is covered by the OP solution that is not in the same way covered by your competition?2020-10-14T10:35:57+02:00

– Cognitive match function that uses unstructured data to create a unique profile per customer or lead (known/unknown)
– Automated feature engineering and machine learning (ML) models based on 10+ years of experience generating conversions in marketing
– Full automation of ML training, scoring, and maintenance processes
– Easy and generic data integration to the customers existing MarTech landscape.
– UI developed for MarTech users.

What value props are your business based on for OP?2020-10-27T09:42:15+01:00

One Prediction makes it simple, fast, and cost-effective to use Machine Learning propensity scores in existing Marketing systems.

– State of the art predictions
– Utilizing existing resources at the client (people, data and systems)
– Proven results within days/weeks
– “Try before you buy”, starting with minimal viable product(s) (MVP)

What unserved needs have you identified in the market that OP addresses better than the competition?2020-10-14T10:35:19+02:00

We are addressing two types of needs:

#1 For the marketing team: Crossing the barrier of rule based communication using ML predictions to personalization and doing it in a way that is scalable, with short time to market, and with no need for data scientist resources.
The result will be increased conversions and ROI.

#2 For the data scientist team: To augment data scientists output by automating and scaling the production of ML predictions for sales and marketing.
This allows data scientists to focus on more advanced tasks related to improving the business.

What are the benefit for a data scientist in One Prediction?2020-10-14T10:41:01+02:00

The need for Machine Learning models will grow over time and there will be a need for many models. This requires an AutoML platform to ensure time-to-market, keep models up to date and minimize resources for day-to-day operations.
One Prediction automates a number of the time-consuming tasks a data scientist usually performs. Your data scientists will have more time to be creative and find new applications for Machine Learning models to optimize your business.
thanks

Our data scientists use different systems. One uses Phyton another R. Question is will One Prediction also be where we should operate and make our models in the future: also the ones we use for TM, paywall etc.?2020-10-14T10:44:12+02:00

One Prediction makes models based on the choice of parameters in marketing language based on the possibilities data provides for models and the model types One Prediction offers. There is no need for coding and you do not have to be a data scientist to make a model.
One Prediction can be used to advantage to create models for all Sales and Marketing activities. Among other things, we have a number of good results on TM activities with other customers.

What is one prediction? What is it not (interfaces other products)? Explain what happens from data is in Bigquery until we can display a piece of content in an email in Portrait. So everyone can understand it.2020-10-14T10:30:47+02:00

One Prediction is a cloud platform that automatically calculates predictive scores within Sales and Marketing, and delivers them to existing systems:
– Orchestration
– DMP + paywall
– TM system
– etc.

How to choose OP between models? is it built-in that it selects random forest if it performs better than linear regression on the individual case?2020-10-14T10:30:19+02:00

One Prediction uses DataRobot as the AutoML component. Watch this video how to choose the best model based on available data (after 30 ”): https://www.youtube.com/watch?v=_JOKlSwKJ10

Is One Prediction a service where you can model any given outcome?2021-02-22T13:58:40+01:00

Yes, you can model any given outcome within Sales and Marketing. One Prediction offers three main types of models:
Response Prediction Models: These models are used to calculate a probability of a given response on a given individual. For example, Buy, Churn, Email Open, PositiveTM response etc. The models can operate at different levels of aggregation. For example, it could be the purchase of a specific product / service or the purchase of a specific category of products / services or just a purchase at all. There is no limit to how many response types can be defined, and an unlimited number of models can be created per. response type at different aggregation levels. For each combination of response type and aggregation level, a Machine Learning model is generated.
Recommendation Models: These models are used to rank a list of products / services, content elements or treatment alternatives per. individual. A treatment is an activity in any in- / outbound touchpoint.
Readiness to Engage Models: These models are used to “time” activities according to the customer’s condition. Example: When is the right time for a travel agency to send an offer for a winter holiday to a specific customer or when is a customer / lead in the market to buy more. The models make a total consideration of the customer’s behavior and calculate probabilities for the individual’s next “move”.

The cognitive match: is it a model you do in One prediction, or is One prediction = cognitive match?2020-10-14T10:29:02+02:00

Cognitive match is a method of incorporating unstructured data in the form of text and images as variables in Machine Learning models. Later, new technologies will make it possible to treat video, sound (music) and other things in the same way.
Cognitive match is only a built-in “feature” in the platform, but still significant and completely unique. When your data about customers / lead, response, content etc. is read in One Prediction, One Prediction will automatically calculate a score on how far an individual and a piece of content match.
Overall, One Prediction is a fully automated Machine Learning cloud platform specifically designed for use by marketers, making it possible to quickly and easily integrate Machine Learning scores into the existing system landscape (Portrait, web, app, Simplify) without large investments, and show results in the form of significantly increased conversion within a few weeks from the start.
You can start by improving a single marketing activity and continuously expand the application, so that you eventually achieve full personalization across all activities and channels.

Based on contact and response history, customer / lead and product data and content elements continuously transferred to One Prediction from data sources (BigQuery and Cxense), One Prediction can fully automate Machine Learning models targeted for use in sales and marketing activities to recommend content, products or offers that best suit the individual. Other models can choose the right audience for specific activities where there is a high probability of positive response from the individual. One Prediction offers a range of standardized models developed on the basis of more than 10 years of experience in marketing automation.

Go to Top