Ten lessons I’ve learned in ten years with Bing

When I joined the Bing team in 2008, our search product was still called Live Search. At that time our market share on the PC was in the single digits—and shrinking steadily.

After much effort over the years and with help from our partners, market share powered by Bing has increased to 35% in the US and over 20% in multiple other markets around the world.[i] Bing is used by half a billion users each month and has become a profitable business, and our team has led innovations that have helped shape the search industry.

It’s both gratifying and humbling to have been part of this journey. So, to honor Bing’s ten-year anniversary, I’d like to share with you ten lessons I’ve picked up along the way.

1. It’s all about the long game

I remember hearing a story from Satya Nadella, who led Bing in its early years. He said that sometimes people plant a tree, and the next day they check on it and wonder why it’s not growing, and then the following day they pull the tree out.

Satya’s story illustrates the impatience that is pervasive in tech and what happens when you invest for the short term. When we launched Bing in 2009, competition was fierce. At the time, only Google was growing market share. We realized that if we wanted to be competitive, we needed to invest deeply and over the long term.

That isn’t easy to do. Many other companies backed off over the years when the growth and returns weren’t evident. But when you believe that something is important, you muster the courage to stick with it through the ups and downs.

Satya, backed by Bill Gates and Steve Ballmer, believed in the importance of our business and the technology we were building. Their long-term investment continues to pay off to this day. Beyond Bing’s own accomplishments, Bing technology has made core contributions to Microsoft products and services including Azure, Office, and Windows.

2. Embrace being the underdog

In the movies, being the underdog can look like fun when you ultimately triumph. But in real life, it’s often hard, and there’s no guarantee of success. People want to be on the winning team. It takes a certain level of internal conviction and humility to admit to being the underdog.

In fact, sometimes it might even be a bit embarrassing. One time, I was watching John Oliver’s show on HBO with my kids, and he poked fun at Bing. My kids were outraged on my behalf and cried out, “Do something! Talk to John!” But the lesson here is that, as the underdog, you must be willing to take occasional jabs and keep pressing on. It’s simply part of the journey.

3. Be metrics driven

One of the great things about being a search engine is the vast amounts of aggregated, anonymized data that can inform the decisions we make and take some of the subjectivity out of the process.

A great example is increasing the length of the titles in our search ads to make them more descriptive. From an aesthetic standpoint, we weren’t sure it was a good choice. But we ran a controlled A/B test and realized that by making this one simple change, we increased user engagement and our revenue by over 10 percent.

We have designed many metrics as proxies for user experience, and being metrics driven has led to better decisions around the features we release. It has also democratized our decisions and helped us create a culture of empowerment and customer obsession. If a new feature or design change improves the user metrics, it ships in the product, regardless of whether the idea came from a VP or an entry-level engineer.

4. But metrics are not everything

Even though our approach is primarily metrics driven and we put much effort into designing the metrics, sometimes they don’t capture the whole story.

Several years ago, major websites began removing the underlines from search results and web experiences. When we tested text without underlines, our metrics suggested that this change was not good. Yet, more and more websites were doing this successfully. We eventually removed underlines based on other measurements and our design team’s guidance, but it was a controversial decision from a metrics point of view. After we made the change, the feedback from our users was clearly positive and they wondered why we didn’t do it sooner.

Looking back, we learned a valuable lesson. While it’s crucial to be metrics driven, it is also important to consider the metric gaps when making ship decisions.

5. Agility is vital

As an online service, Bing was the first product at Microsoft to adopt a more agile approach to shipping features. Product updates used to happen at a much slower pace. We would plan, develop, test, and ship—a release cycle that could take months or even years.

We knew that we needed to speed things up, so we built experimentation and development platforms that allowed us to test, learn, iterate, and ship features quickly. Today we run more than 1,200 experiments each month and release features and updates several times a day. 

This agility has been vital to helping us improve the quality of Bing rapidly over the years. In recent years, other Microsoft teams like Office and Windows have been progressively adopting this more agile approach, which is becoming the norm across the company.

6. Seek technology breakthroughs

One of the benefits in our company is that our engineers can explore and incubate new ideas working closely with world-renowned researchers in Microsoft Research. 

From the early days, Bing embraced the most advanced machine learning technology from Microsoft Research such as FastRank, which accelerated our ranking experimentation by an order of magnitude, and in recent years we have continued to collaborate on deep and reinforcement learning. As an example, Bing was the first general search engine with which users could search and shop for objects within images, leveraging object recognition powered by deep learning.

Seeking breakthroughs is a must in this space, as it is in most high-tech areas, but make sure to bet on technologies that solve a customer problem and avoid those that make a great demo but are not robust or are technologies in search of a problem.

7. Growth is a separate discipline

The traditional way to grow a product is to launch, market, and drive adoption. In the world of internet, we have learned that there are other ways, and we need to think of growth as a separate discipline within our organizations.

Through controlled experiments, we can determine which features are engaging and retaining users. We found that sometimes a simple UX change can significantly increase user engagement. A good email strategy can re-engage users. A new entry point that makes search more convenient can bring in new users. And even small tactics like protecting search preferences can make a large impact on growth and revenue.  

By focusing a core team on growth tactics across multiple features, we have been able to significantly grow Bing, even with headwinds.

8. Trust in search is more important than ever

Twenty years ago, the internet was idealistic. It was a place where people could freely share information in an altruistic way. Unfortunately, over the last several years, we’ve seen the emergence of bad actors who are trying to spread misinformation and fake news.

Here at Bing, we are continually seeking out intelligent ways to serve up trustworthy search engine results and balanced perspectives from reputable sources. And we will continue to do so. It’s critical that all search engines step up, acknowledge the challenge, and find ways to deliver results that are as authoritative and unbiased as possible.

9. Foster a culture of empowerment

I feel so lucky to have been able to work with such amazing people during my time with Bing. Building a culture and nurturing an environment in which people feel empowered and invested has been key to our progress. As mentioned earlier, one of the ways we do this is through our metrics-driven approach and experimentation, where anyone can have the next great idea.

Another way we do this is through collaborative leadership. A great business cannot be built without great people. As a leader, I’m convinced that truly being open to ideas within my entire organization has helped build a culture of mutual respect. The goal is for people to feel safe and empowered to innovate and do their best work.

10. Healthy competition leads to innovation

We at Bing fully embrace the good that competition delivers. In fact, competition is a key reason why I’ve worked at Bing longer than anywhere else.

When I first started at Bing, I worked from our London office and often met with partners including SEO professionals and advertising companies. The message I heard from them time and again was that they were desperate for competition in the search engine space. Competition in search makes Bing better and it makes our competitors’ products better. It leads to more innovation and better pricing for the industry. 

For example, over a year ago we developed rich, immersive experiences for eSports searches in Bing that were way ahead of our competitors. This was great for our users. In recent months, some of our competitors are closing the gap, which is good for their users and is also pressuring us to innovate further. Sometimes we are the ones closing the gap in some query segment and putting pressure on the competition. This is the dynamic that powers state-of-the-art advancements for society, and it is ultimately what matters most in what we do.

Looking back over the last ten years, while my journey with Bing has not always been easy, it has been fulfilling in terms of what we’ve accomplished as a team for the company, and more importantly, for our customers and for society. Although we’ve come a long way, the competition is formidable, and we still have a long way to go. Search is still young and continues to evolve, rapidly becoming more conversational, more visual, and much more natural to engage with. I look forward to what lies ahead in the next ten years.

[i] Bing Network PC Search Share, comScore qSearch, (custom), Dec 2018

Jordi Ribas is a Corporate Vice President at Microsoft in the AI & Research division. He is responsible for Bing product and growth, as well as multiple teams that contribute search technology and intelligence to Azure Cognitive Services, Windows and Office.

Feel free to connect/follow him on LinkedIn and engage on Twitter @JordiRib1