Get the most accurate size recommendation from a video and/or body shape questions.
97% of fashion eCommerce users don't buy and up to 50% of those who buy return their order.
With Presize this changes and net revenue increases by 10% on average due to size recommendations users trust (more conversion) and keep (less returns).
Our solution leads to an 10% increase in net revenue on average (vs. control group with no sizing solution/competing solution):
Increase your conversion rate by up to 25% (relative, compared to a control group)
Reduce your return rate by up to 5% (relative, compared to a control group)
Our AI-algorithm learns from returns of other users with the same body measurements and increases accuracy by +20% in 6-12 months.
Based on a benchmarking study conducted with 255 participants, Presize showed greater body measurement accuracy for 90% of all subjects and scored a mean average error that was 55% lower than that of the second most accurate solution in the benchmarking. You can read the full accuracy evaluation study here.
Users do not always scan themselves under perfect conditions: sometimes the lighting is poor, they are dressed in non-optimal clothing, or they simply do not follow the scanning instructions. Our technology is built to overcome these inconsistencies to deliver the same results, regardless of whether a user's clothing or posture changes. Our competitors fail at this challenge and produce varying measurements for the same person if the scanning conditions diverge from the intended conditions.
Many sizing companies aim to provide size recommendations using a questionnaire-based approach, asking questions such as ‘Does your body look like picture A, B, or C?’ We investigated the efficacy of questionnaire solutions in the context of how well they advise users on their best fit and through A/B testing found that video-based approaches showed far superior results. Body shape questionnaires encourage highly subjective self-reported responses – if users respond at all – 71% of shoppers would answer ‘I don’t know’ if given the option to while 41% of answers are objectively false (e.g. if you compare body mass index with how they perceive their belly). Creating 3D models is also impossible if opting for a questionnaire approach. We do, however, acknowledge that some users may have a preference for a non-video sizing option. Therefore, Presize also offers a user-tested questionnaire solution that produces satisfactory size recommendations.
We automatically scrape all sizing tables from the brand's website and ingest them. Product specific data is also scraped from the specific product detail page and ingested automatically. Presize can combine multiple data sources to deliver high quality size recommendations.
We can scrape product specific data from the product detail page without any effort on your part. Providing additional product or sizing data is completely optional.
We also try to leverage transactional data (returned/unreturned, return reason), when available, in order to accelerate the performance of our size recommendation algorithm. That means we connect the body profile of a user (chest circumference, arm length, etc.) with their behaviour (ordered, returned, etc.) to improve the recommendations given to future visitors with similar bodies. Details on how that works here.
Our solution can be integrated with zero IT capacity and automated data ingestion. We automatically scrape all sizing tables and product data from the brand's website. Since Presize is a White label solution, we can support any corporate branding and are fully GDPR compliant.
Learn more