Discover gfknewron Predict: A real-world, AI-enhanced forecasting solution
![Download full white paper as PDF](https://no-cache.hubspot.com/cta/default/2405078/b7d2f510-b2a7-45be-8470-13a96ea5594f.png)
gfknewron Predict combines the latest AI-led technology with proprietary data sourced by GfK — a giant in market and consumer intelligence since 1934 — to deliver clear insight into the future of product demand and Tech & Durables market activity. Meanwhile, Tactical Forecasting — a tailored solution for retailers powered by gfknewron Predict — gives retailers specialized insights that help optimize supply chains, uncover sales projections per product segment, plan management and promotions, and actively manage sales targets.
Imagine tapping into precise and timely projections of future demand as a retailer or manufacturer, and being able to anticipate what goods customers will buy, where they will buy them, in which quantities, and at what price? Let’s explore gfknewron Predict Forecasting and Retail’s Tactical Forecasting powered by gfknewron Predict to illustrate how it’s possible.
Starting from the dashboard overview, users can drill down and uncover specific insights most relevant to their project and business goal via the channels selection drop-down menu and filters.
Our Customer Success Managers provide timely annotations directly within the forecasts, facilitating near-real-time updates on the client's interface. This seamless integration ensures that market expertise is readily transmitted across the platform, accessible 24/7 to users.
These channel selection, feature and display options ensure gfknewron Predict users can quickly surface and act on the insights that matter. A download option (found on the top right hand corner of each chart) further enables users to save and share their forecasts with colleagues, as an .xls file.
<div class="infogram-embed" data-id="_/o741f3Gdh56eMJZq6NNA" data-type="interactive" data-title="GfK Newron Predict: Infogram1"></div>
Users can choose to display forecasts as absolute or running totals, predicting movement in units or revenue over defined periods. Here, the dashboard displays a 26-week projection based on the most recent week’s proprietary GfK POS data, further enriched with historical data. Users can toggle between display options: switching focus between sales units and revenue at market level, across digital and brick-and-mortar sales.
Updated weekly to reflect insights based on the latest data, this forecast is best used to inform strategic planning that maximizes sales and streamlines inventory and stock management.
Businesses wanting to support longer-term planning also have the advantage of extended, 18-month forecasts. This allows the user to forecast well into the next year, delivering enhanced Full Year planning.
While these Absolute charts better enable benchmarking and a thorough comparative analysis between categories or segments, Running Total charts will also be available on 18-month forecasts. These are tailor-made to provide deeper insight into trends over time and help teams better understand seasonal fluctuations.
Further, users will also quickly be able to compare the past years’ performance for the forecasted period on both charts, with the enrichment data displayed directly below their main forecast for quick visual comparison.
<div class="infogram-embed" data-id="_/yunng3qbPbL68Y6L2pwu" data-type="interactive" data-title="Forecasting Whitepaper Design 7"></div>
gfknewron Predict bases its forecasts on the world’s largest sample of POS data — totaling trillions of unique transactions. Each forecast is generated by a machine learning algorithm that learns from more than 10 years of this data, categorized by product group, product features, region, and distribution. This amounts to nearly 100,000 unique time series. While we capture the long term trends and behaviour and learn from many historical examples, the short term behaviour undoubtedly has a major impact on short term swings.
A bottom-up methodology initially analyzes individual sales over a wide range of time-series data, before aggregating it to segment and market level data to uncover which direction the market is moving in. Informed by external economic data and expert forecasts as well, the gfknewron Predict model delivers forecasts with up to 98% accuracy. Accounting for correlations and complex patterns across industries and locations worldwide, this model unlocks a unique insight into the dynamics and trends of the global marketplace.
Why choose gfknewron Predict as your business forecasting tool?
The world's largest proprietary point of sales database combined with external economic data and a unique holistic forecasting model to deliver unprecedented insights into the fast changing retail markets worldwide.
We test all our forecasts against a metric called (1-RMSSE). In simple terms, this compares the outperformance of our forecasts against simply using a baseline forecast from the prior year. This ensures that our forecasts scientifically outperform and add value against a direct alternative. This corroborates with qualitative feedback from clients that our forecasts lead the market in terms of accuracy.
The proprietary POS data set includes key metrics, such as sales volume, revenue, and individual store information. Additional calendar data for significant events — such as the Blavatnik School of Government’s COVID-19 Response Tracker to ensure historical data during COVID times can be normalized and used — complete the data set. These are analyzed in relation to sales trends and macroeconomic factors including regional GDP, consumer confidence and employment rates, ensuring the latest economic trends are also captured.
Most importantly, the gfknewron Predict model is engineered to constantly assess and incorporate new data sources as they become available — ensuring continued accuracy and relevance. Further, these insights (data and recommendations) are presented in clear, customizable, and easy-to-understand dashboards — enabling teams across organizations to self-serve their scenario planning needs and base their decisions on high-quality forecasts.