Right Image R&D
6 likes1,373 views
Right Image R&D is developing AI to optimize mobile app marketing visuals like icons and screenshots by maximizing click-through rates. Their solution would automatically test different images and learn which perform best without slow A/B testing. This has applications for mobile app promotion, a $40 billion annual market concentrated in games. While competitors offer self-serve A/B testing or paid clickers, Right Image R&D's advantage is automation through AI for quick, low-cost results. They plan to launch with direct sales and agencies, starting with proof of concept in mid-2016 and first AI customer by September 2016.
Right Image R&D
- 1. Right Image R&D AI that maximizes CTR of your marketing visuals April 2016
- 2. 1. Problem: slow A/B testing Context: suboptimal marketing content leads to lost sales +41% -51%
- 3. Our successful manual A/B testing case: 6x increase, lasting impact Mobile Game. Genre: Strategy. Start of app store page optimization Monthly installs by device +480% Lasting impact achieved: 5-6x average monthly installs lift
- 4. 2. Solution: instant image ranking Applications: o Mobile app icon & screenshots o Ad banners etc.
- 5. 3. Market: mobile app promotion o Mobile app market: –$40 billion / year –5-year CAGR: 20% –Mobile app promotion market (est.): $4-8B o Concentrated: –85% of revenue comes from games –2% of app developers claim 54% of all app revenues Source: App Annie Report: http://files.appannie.com.s3.amazonaws.com/reports/App-Annie-02-2016-Forecast-EN.pdf?aliId=80125619 http://www.techrepublic.com/article/mobile-developers-should-spend-less-time-on-development-more-on-marketing/
- 6. 4. Competition o Self-serve A/B testing platforms o A/B testing platforms that provide clickers o RightImage R&D competitive advantage: –Automation, enabled by AI –Quick results at low cost
- 7. 5. How it works: learn from SM images
- 8. Apply analysis to marketing visuals
- 9. Usage-based, to be decided: ● Charging for every target action we’ve added ● Charging for volume of tested pictures 6. Business Model
- 10. 7. Marketing o Direct sales to massive app owners o SEO & landing page (planned) o Affiliated marketing agencies
- 11. 8. Roadmap Now: 1 served customer July 2016: 5-10 manually served customers July 2016: AI proof of concept Sep 2016: first customer served with AI
- 12. 9. Team Michael Rybak Strategy, R&D Interests ● Human perception & behavior ● Algorithms & optimization Experience ● 3 years as an R&D team lead ● 12 years as research engineer ● 2 years as a project manager ● 2 years as CEO & Coach of a soft skills trainings startup Education ● 3 postgraduate programs in management & leadership ● Red at TopCoder ● Bronze medal at ACM’2003 ● Bronze medal at IOI’2001 ● M. Sc. in Computer Science, Kyiv Shevchenko Univ ’2007 Mykola Komarevskyy Strategy, Business Development Denis Troyanov R&D Interests ● Deep learning & computer vision, neuroscience & AI Experience ● Video stream analysis ○ fire-detection system ○ for warehouse safety ○ commercially applied ● Computer vision for audience analysis: ○ 20%+ precision raise of face-processing technology core of a running business Education ● Computer vision school ● B.Sc. in Computer Science Interests ● Technology in business and society, machine learning Experience ● 7+ years of leadership in software (cofounder, business development, CTO, product) ● 2+ years at McKinsey & Co., Business Technology Education ● MBA, Thunderbird School of Global Management, USA ● B.Sc. in Computer Science, Shevchenko University of Kyiv ● Coursera courses: Machine Learning, Big Data, IoT
- 13. Your feedback & ideas ● Ecosystem ○ Products? ○ Technologies? ○ Other applications & markets of image sentiment ranking? ● Partnership ○ Interested to become our beta client? ○ Interested to become a partner (marketing agency)? ○ Know someone who could be interested? ● Team ○ Known ML specialists?