The herd sours on unprofitable unicorns again

There has been a mountain of press lately about how investors are souring on unprofitable unicorns.

We’ve seen this movie before; for a while, it’s all about growth, and profits be damned, then the winds change, and everyone focuses on “capital efficiency,” or similar jargon meaning, “how can I get big returns without having to put much money at risk?”

The winds blow back and forth. Until very recently, everyone was in love with consumer unicorns again. Now investors are licking their wounds, except for those who eschewed the name brands and went for boring old B2B and infrastructure companies. They are doing just fine, thank you.

Why are investors overpaying for household-name unicorns? Is it that they really believe they are good assets, or are other factors at play? The fact is that venture funds and private equity funds are competing for investment funds themselves. I am personally an investor in several venture funds, and I have heard the pitch, “we were investors in Facebook, Instagram, Uber, Twitter (or whatever), and we can get you access to these deals.” Sounds good, but what they don’t tell you is how much they paid (or overpaid) to be part of these deals. It’s such nonsense and the perennially poor returns delivered by the ego-driven venture capital industry are its just rewards.  

Downward stock valuations of unicorns post-IPO

Downward stock valuations of unicorns post-IPO. (Yahoo Finance)

Investors who bid up the valuations of high-profile unicorns are of course hoping that an IPO will eventually bail them out. The problem is that public fund managers, like Fidelity or Blackstone, who control most of post-IPO stocks, look at the value of a company quite differently. They see a company’s “value” as the sum total of all the company’s future profits. They can’t offer their clients “exclusive” access to hot deals. We’re talking public stocks that anyone can buy.

If nobody can see a clear road to profitability, then this hard-nosed approach to valuation will lead to stocks tanking after an IPO. That’s recently been the case with We, Uber, and numerous others.  

From 2010 to the first quarter of 2015, investors collectively poured $9.4 billion into the on-demand economy, according to data from CB Insights. Uber accounted for 58% of the $4.12 billion raised in 2014. What’s also striking is how quickly the industry piled onto the latest thing between 2013 and 2014. Since its IPO in May of 2019, Uber’s stock has fallen nearly 40% from its peak, Lyft is down even more, and Softbank’s most recent investment in We appears to have wiped out nearly 80% of its previous private valuation. Masayoshi Son has been publicly apologizing for his investment in We: “My investment judgment was poor in many ways,” said Son.

Direct mail still works if you avoid common mistakes

We’ve aggregated many of the world’s best growth marketers into one community. Twice a month, we ask them to share their most effective growth tactics, and we compile them into this Growth Report.

This is how you stay up-to-date on growth marketing tactics — with advice that’s hard to find elsewhere.

Our community consists of 1,000 startup founders and VP’s of growth from later-stage companies. We have 400 YC founders, plus senior marketers from companies including Medium, Docker, Invision, Intuit, Pinterest, Discord, Webflow, Lambda School, Perfect Keto, Typeform, Modern Fertility, Segment, Udemy, Puma, Cameo and Ritual .

You can participate in our community by joining Demand Curve’s marketing webinars, Slack group, or marketing training program.

Without further ado, onto our community’s advice.

Advertising in Discord/Telegram communities

Insights from Varun Mathure of Midnite

Discord/Telegram can be a great place to find engaged, niche communities for advertising. However, do not treat it like a typical ad channel. Community marketing is its own art, and there are many principles to doing it effectively. Here are just a few:

  • Treat Discord/Telegram users like you would Reddit users: they’ll reject being advertised to unless there’s legitimate, authentic value being provided.
  • Work with moderators to offer services that make their moderation duties easier. Perhaps a bot or tool that would be legitimately useful to the community while also organically pitching your startup.
  • Have a well-respected community member vouch for you — it goes a long way toward building trust with the rest of the community. Always start by building relationships.
  • Have a member of your team active in the community. Don’t just advertise; contribute regularly.
  • Run promos/incentives that encourage members to post your product screenshots or share your product output in the community. In other words, incentivize a frictionless way for community members to become your brand ambassadors.

Landing page tear-downs [Video]

Watch us critique landing pages. In the process, you’ll learn how to improve your own.

Most common direct mail mistakes

Define and manage growth on your own terms

Welcome to this edition of The Operators, a recurring Extra Crunch column, podcast, and YouTube show that brings you insights and information from inside top tech companies. Our guests are execs with operational experience at fast-rising startups, like Brex, Calm, DocSend, and Zeus Living, and more established companies, like AirBnB, Facebook, Google, and Uber. Here, they share strategies and tactics for building your first company and charting your career in tech.

In this episode, we’re talking about growth. Growth means different things inside different organizations, but correctly identifying avenues for sustainable and scalable growth is a priority for almost all companies. We’ll cover:

  1. Defining growth and being good at it
  2. Managing growth without losing sight of the big picture
  3. How companies should approach growth

To learn more, we spoke with two experts:

Isaac Silverman began his career as an entrepreneur before joining Zynga to work on growth development. At Zynga, he focused on some of the most cutting-edge approaches to growth and development. He then moved to Postmates, where he focused on growth product and is now the head of rider growth at Uber.

Matias Honorato is a senior manager on the growth team at Tally, a growth-stage tech company, and also brings his own entrepreneurial roots and experience at companies like Earnest and Tradecraft.

Below is a summary of our conversation; check out The Operators for the full episode.

Defining growth and being good at it

Growth as a concept and discipline originates from the term “growth hacking.” It can be hard to grasp as distinct from functions and goals that usually sit with the marketing team or product development team and may be best thought of as a combination of both. We think of it as the domain responsible for designing, implementing, and measuring approaches to acquiring and retaining customers. It’s a mix of marketing and product, but also sales and data analytics, and sometimes even operations.

Great growth professionals can be successful with a wide variety of work or educational backgrounds, and are most often curious, persistent, and adept at thinking holistically, creatively, quantitatively, and interdisciplinarily.

“There’s definitely a lot of deep analysis and how all the pieces fit together and there’s a lot of product work, and there’s a lot of marketing work,” said Silverman. “I think part of what I find so deeply interesting and engaging about it is it brings together everything. It’s really the exercise we go through, and I don’t want to overstate our role, but the exercise we go through is, ‘let’s imagine that we’re the CEO and what are the things that we think are really important. Let’s see the whole picture and then figure out what are the areas that we should ultimately focus on within it.’ So that is ultimately deeply, deeply, stimulating and dynamic and changes on a day to day basis. And sometimes it’s more product manager-y, sometimes it’s more something else.”

Honorato said that to be a great growth professional, “you have to have a really good understanding of your business, what are your goals, how the product works, how their financial side of the business works.”

The responsibilities of growth teams range from simple tasks like split-testing marketing copy and landing pages to more complex strategies like enabling the integration of a file storage and management solution into workflow applications and then subsequently partnering with those workflow applications to acquire users and become a default solution. Being cross-functional in nature, growth initiatives often require resources and contributions from other teams like marketing, design, and engineering. This can create conflict due to resource constraints and company politics, regardless of how small or large a company is. These are meaningful challenges before even evaluating the effectiveness of growth initiatives! Great growth teams must know how to navigate these types of issues as well, making effective growth teams hard to build, but very valuable if you can build an effective one.

“I tend to believe teams exist on spectrum,” said Silverman. “You got that sort of optimizer or specific functionality or specific parts of the funnel or whatever growth themes and then in the spectrum you have, the entire purpose of the company after you’ve achieved product market fit is to grow. I tend to believe that a lot of companies think they need the former and actually need the latter… One thing that I want to make sure is absolutely clear, the growth at Uber is the product of a very high number of very, very competent people, very diligently thinking about their part of the business, and [growth is] a portion of that much, much larger equation.”

Managing growth without losing sight of the big picture

Reimagine inside sales to ramp up B2B customer acquisition

Slack makes customer acquisition look easy.

The day we acquired our first Highspot customer, it was raining hard in Seattle. I was on my way to a startup event when I answered my cell phone and our prospect said, “We’re going with Highspot.” Relief, then excitement, hit me harder than the downpour outside. It was a milestone moment – one that came after a long journey of establishing product-market fit, developing a sustainable competitive advantage, and iterating repeatedly based on prospect feedback. In other words, it was anything but easy.

User-first products are driving rapid company growth in an era where individuals discover, adopt, and share software they like throughout their organizations. This is great if you’re a Slack, Shopify, or Dropbox, but what if your company doesn’t fit that profile?

Product-led growth is a strategy that works for the right technologies, but it’s not the end-all, be-all for B2B customer acquisition. For sophisticated enterprise software platforms designed to drive company-wide value, such as Marketo, ServiceNow and Workday, that value is realized when the product is adopted en masse by one or more large segments.

If you’re selling broad account value, rather than individual user or team value, acquisition boils down to two things: elevating account based-selling and revolutionizing the inside sales model. Done correctly, you lay a foundation capable of doubling revenue growth year-over-year, 95 percent company-wide retention, and more than 100 percent growth in new customer logos annually. Here are the steps you can take to build a model that realizes on-par results.

Work the account, not the deal

Account-based selling is not a new concept, but the availability of data today changes the game. Advanced analytics enable teams to develop comprehensive and personalized approaches that meet modern customers’ heightened expectations. And when 77 percent of business buyers feel that technology has significantly changed how companies should interact with them, you have no choice but to deliver.

Despite the multitude of products created to help sellers be more productive and personal, billions of cookie-cutter emails are still flooding the inboxes of a few decision makers. The market is loud. Competition is cut throat. It’s no wonder 40 percent of sales reps say getting a response from a prospect is more difficult than ever before. Even pioneers of sales engagement are recognizing the need for evolution – yesterday’s one-size-fits-all approach to outreach only widens the gap between today’s sellers and buyers.

Companies must radically change their approach to account-based selling by building trusted relationships over time from the first-touch onward. This requires that your entire sales force – from account development representatives to your head of sales – adds tailored, tangible value at every stage of the journey. Modern buyers don’t want to be sold. They want to be advised. But the majority of companies are still missing the mark, favoring spray-and-pray tactics over personalized guidance.

One reason spamming remains prevalent, despite growing awareness of the need for quality over quantity, is that implementing a tailored approach is hard work. However, companies can make great strides by doing just three things:

  • Invest in personalization: Sales reps have quota, and sales leaders carry revenue targets. The pressure is as real as the numbers. But high velocity outreach tactics simply don’t work consistently. New research from Monetate and WBR Research found that 93% of businesses with advanced personalization strategies increased their revenue last year. And while scaling personalization may sound like an oxymoron, we now have artificial intelligence (AI) technology capable of doing just that. Of course, not all AI is created equal, so take the time to discern AI-powered platforms that deliver real value from the imposters. With a little research, you’ll find sales tools that discard  rinse-and-repeat prospecting methods in favor of intelligent guidance and actionable analytics.

Know your startup’s value so you can communicate it to investors

I’ve always told companies that investors have a much easier job than they do. To be good at their jobs, investors have to know how to do math and make decisions. As a business owner, you have to do both while also running your business.

The math piece can seem cumbersome, but it’s vital for understanding whether your company is creating or destroying value. A few simple metrics can demonstrate to investors the health and viability of your company, and they can show you which levers to pull that will best optimize your company for investor interest (and secure a higher price). But before you can ever hope to communicate your business’ value to an investor, you must understand it yourself.

The numbers are simple; it’s the calculations that are complex

Investment math itself is not complicated. In essence, it’s just about understanding whether your company is creating or destroying value by asking:

  • Where is your company investing its financial resourcesMost growing companies invest heavily in sales and marketing or research and development.
  • What is the return on this investment?  For example, how much gross profit (revenue x gross margin percentage) does a given sales and marketing investment produce?
  • How does that number compare to your cost of capital? If it’s higher, your company is creating value. If it’s lower, you’re destroying it.

Investors use this information to determine if their return would be higher than their expectation (e.g., 15% hurdle rate), should you continue down your current path of creating or destroying value. Then, they make their decision based on that calculation.

A caveat I’ll add here is that it’s not necessarily a deal-breaker if your company is declining in value. Oil rigs, after all, are considered investment assets, even though they are perpetually declining and will eventually run out (i.e., destroy all of their value). Although this article focuses on calculations that demonstrate value creation, all investment assets can be financed at the right price.

A deep dive into calculating value

One of the best metrics you can use to demonstrate value creation is your cohort-level return on investment. It’s a calculation most investors are familiar with, but it may not be as straightforward to companies who don’t see it as often. Again, while the metrics and concepts of investment math are simple, it’s the process of getting there that requires complex analysis.

Whether you are evaluating these metrics yourself or bringing in outside counsel to assist you, use the process below to show investors you are creating value.

Determine which information to analyze

The first step in calculating value is to understand which information from your income and cash flow statements to analyze as “investments.”

Start by dividing your capital allocation into three main buckets: short-term investments, long-term investments and expenses. In general, short-term investments will be the ones you want to focus on, but it’s helpful to walk through each.

  • Short-term investments (pay back within 24 months)

Decide which type of investor to target for raising capital

I recently wrote Should you raise venture capital from a traditional equity VC or a Revenue-Based Investing VC? Since then, I’ve talked with a number of other firms and greatly expanded my database: Who are the major Revenue-Based (RBI) Investing VCs?

That said, venture capital is just one of many options to finance your business, typically the most expensive. The broader question is, what type of capital should you raise, and from whom?  

I find many CEOs/CFOs default to approaching investors who have the most social media followers; who have spent the most money sponsoring events; or whom they met at an event. But, fame and the chance that you met someone at a conference do not logically predict that investor is the optimal investor for you. In addition, the best-known investors are also the ones who are most difficult to raise capital from, precisely because they get the most inbound.

The first step is to decide the right capital structure for your financing. Most CFOs build an Excel model and do a rough comparison of the different options. Some firms provide tools to do this online, e.g., Capital’s Cost of Equity estimator; Lighter Capital’s Cost of Capital Calculator; 645 Ventures’ cap table simulator. A similar, open-source, highly visual tool focused on VC is Venture Dealr.

For each of the major categories of investors, you can find online databases of the major providers. Major options include:

  • Traditional equity venture capital and private equity. For early-stage startups in particular, I suggest Foundersuite*, Samir Kaji’s Master List of US Micro-VC’s and Shai Goldman’s database of VC funds at/below $200M in size. You can find other databases of investors at AngelList, CB Insights, Crunchbase, Dow Jones VentureSource, Pitchbook, Preqin, and Refinitiv Eikon
  • Revenue-based investing VC. See Who are the major Revenue-Based Investing VCs?
  • Venture debt. See FindVentureDebt and this comparison guide of debt options for SAAS companies. Watch out for double dipping, or interest on interest.
  • Merchant cash advances/factoring. See Debanked’s list.
  • Small Business Association Loans. Ravi Bhagavan, Managing Director, BRG Capital Advisors, said, “a low-cost and often convenient form of capital for small businesses is SBA loans, which are guaranteed by the Small Business Administration. SBA loans are $5k – $5M in size and are typically at a lower cost of capital compared to alternate forms of debt, since up to 85% of the loan is guaranteed by the SBA. Additionally, SBA loans have longer payment periods (5-25 years) than traditional forms of financing and come with less onerous ongoing disclosure requirements. However, SBA loans typically require a personal guarantee (PG) from the founder(s), who are scrutinized for income and credit history at the time of application. PGs can be quite daunting to founders because it puts their personal assets, including homes and investment accounts, on the line. SBA loans are available through SBA-approved banks and SBIC funds. SBICs make equity and debt investments of size $100k – $10M in qualifying small businesses. A good resource for looking up SBICs is here.” 
  • Crowdfunding, e.g., Republic*, Indiegogo*.  This option provides you capital and also market validation for desire for your product.  

Once you decide on the right category of investor, here are some tools I suggest using to find the optimal capital provider:

  • Most important, reference checking. I have a whitelist of investors I recommend to my portfolio — and a blacklist which I guide them to avoid.
  • Comparison websites: BitX, Fundera, GUD Capital, Lencred.com, Lendio, and NerdWallet Small Business Loans are all resources which can help you evaluate different options for small business financing, typically within a defined category of financing. Braavo specializes in financing app companies.
  • Financing supermarkets: Most investment firms start out with one asset class, and then over time they often add others. There are countless examples, e.g., most of the large B2B banks, Kapitus, Kalamata Capital, United Capital Source, etc. These firms can give you an apples-to-apples comparison of what different capital forms, albeit all from one provider, will cost you.

California’s new data privacy law brings U.S. closer to GDPR

Data privacy has become one of the defining business and cultural issues of our time.

Companies around the world are scrambling to properly protect their customers’ personal information (PI). However, new regulations have actually shifted the definition of the term, making everything more complicated. With the California Consumer Privacy Act (CCPA) taking effect in January 2020, companies have limited time to get a handle on the customer information they have and how they need to care for it. If they don’t, they not only risk being fined, but also loss of brand reputation and consumer trust — which are immeasurable.

California was one of the first states to provide an express right of privacy in its constitution and the first to pass a data breach notification law, so it was not surprising when state lawmakers in June 2018 passed the CCPA, the nation’s first statewide data privacy law. The CCPA isn’t just a state law — it will become the defacto national standard for the foreseeable future, because the sheer numbers of Californians means most businesses in the country will have to comply. The requirements aren’t insignificant. Companies will have to disclose to California customers what data of theirs has been collected, delete it and stop selling it if the customer requests. The fines could easily add up — $7,500 per violation if intentional, $2,500 for those lacking intent and $750 per affected user in civil damages.

Evolution of personal information

It used to be that the meaning of personally identifiable information (PII) from a legal standpoint was clear — data that can distinguish the identity of an individual. By contrast, the standard for mere PI was lower because there was so much more of it; if PI is a galaxy, PII was the solar system. However, CCPA, and the EU’s General Data Protection Regulation GDPR, which went into effect in 2018, have shifted the definition to include additional types of data that were once fairly benign. The CCPA enshrines personal data rights for consumers, a concept that GDPR first brought into play.

The GDPR states: “Personal data should be as broadly interpreted as possible,” which includes all data associated with an individual, which we call “contextual” information. This includes any information that can “directly or indirectly” identify a person, including real names and screen names, identification numbers, birth date, location data, network addresses, device IDs, and even characteristics that describe the “physical, physiological, genetic, mental, commercial, cultural, or social identity of a person.” This conceivably could include any piece of information about a person that isn’t anonymized.

With the CCPA, the United States is playing catch up to the GDPR and similarly expanding the scope of the definition of personal data. Under the CCPA, personal information is “information that identifies, relates to, describes, is capable of being associated with, or could reasonably be linked, directly or indirectly, with a particular consumer or household.” This includes a host of information that typically don’t raise red flags but which when combined with other data can triangulate to a specific individual like biometric data, browsing history, employment and education data, as well as inferences drawn from any of the relevant information to create a profile “reflecting the consumer’s preferences, characteristics, psychological trends, preferences, predispositions, behavior, attitudes, intelligence, abilities and aptitudes.”

Know the rules, know the data

These regulations aren’t checklist rules; they require big changes to technology and processes, and a rethinking of what data is and how it should be treated. Businesses need to understand what rules apply to them and how to manage their data. Information management has become a business imperative, but most companies lack a clear road map to do it properly. Here are some tips companies can follow to ensure they are meeting the letter and the spirit of the new regulations.

  • Figure out which regulations apply to you

The regulatory landscape is constantly changing with new rules being adopted at a rapid rate.  Every organization needs to know which regulations they need to comply with and understand the distinctions between them. Some core aspects CCPA and GDPR share include data subject rights fulfillment and automated deletion. But there will be differences so having a platform that allows you to handle a heterogenous environment at scale is important.

  • Create a privacy compliance team that works well with others

What AI startups need to achieve before VCs will invest

Funding of artificial intelligence-focused companies reached approximately $9.3 billion in the U.S. in 2018, an amount that will continue to rise as the transformative impact of AI is realized. That said, not every AI startup has what it takes to secure an investment and scale to success.

So, what do venture capitalists look for when considering an investment in an AI company?

What we look for in all startups

Some fundamentals are important in any of our investments, AI or otherwise. First, entrepreneurs need to articulate that they are solving a large and important problem. It may sound strange, but finding the right problem can be more difficult than finding the right solution. Entrepreneurs need to demonstrate that customers will be willing to switch from what they’re currently using and pay for the new solution.

The team must demonstrate their competence in the domain, their functional skills and above all, their persistence and commitment. The best ideas likely won’t succeed if the team isn’t able to execute. Setting and achieving realistic milestones is a good way to keep operators and investors aligned. Successful entrepreneurs need to show why their solution offers superior value to competitors in the market — or, in the minority of cases where there is an unresolved need — why they’re in the best position to solve it.

In addition, the team must clearly explain how their technology works, how it differs and is advantageous relative to existing competitors and must explain to investors how that competitive advantage can be sustained.

For AI entrepreneurs, there are additional factors that must be addressed. Why? It is fairly clear that we’re in the early stages of this burgeoning industry which stands to revolutionize sectors from healthcare to fintech, logistics to transportation and beyond. Standards have not been settled, there is a shortage of personnel, large companies are still struggling with deployment, and much of the talent is concentrated in a few large companies and academic institutions. In addition, there are regulatory challenges that are complex and growing due to the nature of the technology’s evolutionary aspect.

Here are five things we like to see AI entrepreneurs demonstrate before making an investment:

Demonstrate mastery over their data and its value: AI needs big data to succeed. There are two models: companies can either help customers add value to their data or build a data business using AI. In either case, startups must demonstrate that the data is reliable, secure and compliant with all regulatory rules. They must also demonstrate that AI is adding value to their own data — it must explain something, derive an explanation, identify important trends, optimize or otherwise deliver value.

With the sheer abundance of data available for companies to collect today, it’s imperative that startups have an agile infrastructure in place that allows them to store, access and analyze this data efficiently. A data-driven startup must become ever more responsive, proactive and consistent over time.

AI entrepreneurs should know that while machine learning can be applied to many problems, it may not always yield accurate predictions in every situation. Models may fail for a variety of reasons, one of which is inadequate, inconsistent or variable data. Successful mastery of the data demonstrates to customers that the data stream is robust, consistent and that the model can adapt if the data sources change.

Entrepreneurs can better address their customer needs if they can demonstrate a fast, efficient way to normalize and label the data using meta tagging and other techniques.

Remember that transparency is a virtue: There is an increased need in certain industries — such as financial services — to explain to regulators how the sausage is  made, so to speak. As a result, entrepreneurs must be able to demonstrate explainability to show how the model arrived at the result (for example, a credit score). This brings us to an additional issue about accounting for bias in models and, here again, the entrepreneur must show the ability to detect and correct bias as soon as they are found.

The AI stack that’s changing retail personalization

Consumer expectations are higher than ever as a new generation of shoppers look to shop for experiences rather than commodities. They expect instant and highly-tailored (pun intended?) customer service and recommendations across any retail channel.

To be forward-looking, brands and retailers are turning to startups in image recognition and machine learning to know, at a very deep level, what each consumer’s current context and personal preferences are and how they evolve. But while brands and retailers are sitting on enormous amounts of data, only a handful are actually leveraging it to its full potential.

To provide hyper-personalization in real time, a brand needs a deep understanding of its products and customer data. Imagine a case where a shopper is browsing the website for an edgy dress and the brand can recognize the shopper’s context and preference in other features like style, fit, occasion, color etc., then use this information implicitly while fetching similar dresses for the user.

Another situation is where the shopper searches for clothes inspired by their favorite fashion bloggers or Instagram influencers using images in place of text search. This would shorten product discovery time and help the brand build a hyper-personalized experience which the customer then rewards with loyalty.

With the sheer amount of products being sold online, shoppers primarily discover products through category or search-based navigation. However, inconsistencies in product metadata created by vendors or merchandisers lead to poor recall of products and broken search experiences. This is where image recognition and machine learning can deeply analyze enormous data sets and a vast assortment of visual features that exist in a product to automatically extract labels from the product images and improve the accuracy of search results. 

Why is image recognition better than ever before?

retail and artificial intelligence

 

While computer vision has been around for decades, it has recently become more powerful, thanks to the rise of deep neural networks. Traditional vision techniques laid the foundation for learning edges, corners, colors and objects from input images but it required human engineering of the features to be looked at in the images. Also, the traditional algorithms found it difficult to cope up with the changes in illumination, viewpoint, scale, image quality, etc.

Deep learning, on the other hand, takes in massive training data and more computation power and delivers the horsepower to extract features from unstructured data sets and learn without human intervention. Inspired by the biological structure of the human brain, deep learning uses neural networks to analyze patterns and find correlations in unstructured data such as images, audio, video and text. DNNs are at the heart of today’s AI resurgence as they allow more complex problems to be tackled and solved with higher accuracy and less cumbersome fine-tuning.

How much training data do you need?

The AI stack that’s changing retail personalization

Consumer expectations are higher than ever as a new generation of shoppers look to shop for experiences rather than commodities. They expect instant and highly-tailored (pun intended?) customer service and recommendations across any retail channel.

To be forward-looking, brands and retailers are turning to startups in image recognition and machine learning to know, at a very deep level, what each consumer’s current context and personal preferences are and how they evolve. But while brands and retailers are sitting on enormous amounts of data, only a handful are actually leveraging it to its full potential.

To provide hyper-personalization in real time, a brand needs a deep understanding of its products and customer data. Imagine a case where a shopper is browsing the website for an edgy dress and the brand can recognize the shopper’s context and preference in other features like style, fit, occasion, color etc., then use this information implicitly while fetching similar dresses for the user.

Another situation is where the shopper searches for clothes inspired by their favorite fashion bloggers or Instagram influencers using images in place of text search. This would shorten product discovery time and help the brand build a hyper-personalized experience which the customer then rewards with loyalty.

With the sheer amount of products being sold online, shoppers primarily discover products through category or search-based navigation. However, inconsistencies in product metadata created by vendors or merchandisers lead to poor recall of products and broken search experiences. This is where image recognition and machine learning can deeply analyze enormous data sets and a vast assortment of visual features that exist in a product to automatically extract labels from the product images and improve the accuracy of search results. 

Why is image recognition better than ever before?

retail and artificial intelligence

 

While computer vision has been around for decades, it has recently become more powerful, thanks to the rise of deep neural networks. Traditional vision techniques laid the foundation for learning edges, corners, colors and objects from input images but it required human engineering of the features to be looked at in the images. Also, the traditional algorithms found it difficult to cope up with the changes in illumination, viewpoint, scale, image quality, etc.

Deep learning, on the other hand, takes in massive training data and more computation power and delivers the horsepower to extract features from unstructured data sets and learn without human intervention. Inspired by the biological structure of the human brain, deep learning uses neural networks to analyze patterns and find correlations in unstructured data such as images, audio, video and text. DNNs are at the heart of today’s AI resurgence as they allow more complex problems to be tackled and solved with higher accuracy and less cumbersome fine-tuning.

How much training data do you need?