Built for Speed eCommerce Imperative: Tips to increase the agility of your business

“The pace of change has never been this fast, yet it will never be this slow again.”

For eCommerce, 2020 was a challenging but booming year. Every country and most industries saw significant growth in their online sales and there is no substantial pull-back in sight as there are more and more new users opting to use the digital channels. Unfortunately, the latest growth is also the biggest trap in this sector. In such times of growth, amidst plenty of success stories in online sales, it is easy to be complacent and don’t feel the need to transform. Meanwhile, there is now an army of new competitors in the market - Several top brands are now going direct-to-customer, even more traditionally offline retailers made big strides towards online sales, development agencies have been on a hiring frenzy to cope with the high demand for innovation and new projects, all of this while marketplaces are getting a bigger share of the overall market. In the coming years, when growth levels normalize, the competition landscape will be tough and many ecommerce operations will struggle to survive. With an increasingly demanding customer, additional external competition, and technology changes it will be the case that only those who can quickly adapt will be able to succeed consistently. It is the new imperative: Evolve fast or die slowly. 

So, How to create an organization that changes fast?  

Having experienced 17 years working with digital technologies, I had the opportunity to live in 6 different countries and across various industries which gave me a wonderful range of organizations to understand some patterns of what makes an organization agile. I would like to share here some of the key tips and insights to make your organization evolve faster - I hope it will be helpful for your future journey. 

People & organizations

This is the biggest factor and also the hardest part. In top companies nowadays there is a hunger to hire and promote people with a digital, data-first mentality that are comfortable with technologies and technical concepts, and are able to create architectures and drive the transformation mindset needed to be a fast company. Such a profile is in drastic shortage - which is the key challenge for most organizations. Here are some tips on ways to build your organizations:

1. Get digital talent, NOW - it is already hard to find, but chances are that it will become increasingly difficult as most companies have been catapulted to digital transformations and most are lagging on digital talent. 
2. As a general rule, favor digital talent over deep industry experience - While industry experience is extremely important if not critical at times, today HR processes are far too biased in screening industry experience and most recruiting processes are not giving digital talent the proper importance. 
3. Cultivate digital talent - It is so hard to recruit top digital talent that leading companies already built capabilities to nurture their own digital leaders. The recommendation here is to identify them and work to see what is blocking them in transforming the organization. Set clear plans to develop these people as this will make all the difference in the years to come. It is important to mention here that I am not talking about “digital natives” vs older generation. While newer generations have an advantage being more fluent in the digital world, digital talent is not about knowing how to use technology, it is more about understanding technological concepts and being able to apply technology toward a goal - this talent is cross generational. Find them and develop them. 
4. Consider the creation of a horizontal innovation taskforce - this applies to the more complex organizations, where creating a team of cross function/department volunteers that are tasked to represent their areas on innovation projects. It is very important that this team is composed of volunteers as the role of innovators should not be assigned to people that do have the initiative and will to participate.
5. Empower the lower levels of the organization - During the initial months of the pandemic I talked to a few ecommerce directors in expectation that, as the online was the only available channel, their influence on the organization would have significantly increased. It was surprising to learn that in some cases it was exactly the contrary. One manager told me that their team could not even send out a newsletter for months, since it did not have the latest campaign image, or the wrong product name, etc. It was clear that in these companies senior management effectively blocked the agility of the online team and directly hurt sales. On the other hand, I saw a fast company who during the peak of the confinement, gave full authority to a marketing associate to communicate with the customers. This allowed fast messages about closures, changes of delivery estimates, messages of home office products, and a far more human communication with the clients in a time that it was needed. It is critical that in times of very fast changes that the lower levels are able to make changes as autonomously as possible. On larger organizations we often see central global organizations that are usually born as subject matter experts, that with time evolved into powerful centralized control groups. This centralization often makes the cost of policies and technology stack lower, but can drastically reduce innovation speed.  
6. Implement one of many agile methodology (e.g. SCRUM, Lean Development, Crystal, Kanban, etc) - There are several passionate discussions out there as to which is the better one. In my personal experience, they are all equally good if properly implemented. Focus on implementing any of them… but implement it well. 
7. Test, Test, Test.. and if all tests fail… keep testing! Google coined a term to make clear to all it’s employees of the traditional managerial approach that should NOT be followed. They would instruct to always avoid the HIPPO (“HIghest Paid Person Opinion'') approach. Far too often we see in companies the HIPPO approach where the top manager ends up taking a decision based on gut feel. Leading companies do not count on one person’s opinion, they create hypotheses and test everything. eCommerces like Amazon, Netflix, Booking.com, Shopify, and many other leaders are known to execute thousands of A/B tests a month.  

Business Model & Operations

Fundamentally business has to be efficient and make money - if i can sell for more than it costs me I have a business. Then the business grows.. We get departments, integrated systems and ever evolving processes, etc. As businesses grow, the mentality of efficiency is always present. The opportunity that we have in all businesses is that as businesses grow more complex, the management of efficiency tends to become more and more transactional. They focus on a few metrics that rarely tell the full picture.  In most companies the “unit cost” is the main objective and not the “system cost”.    

One of the best cases of the change of mentality needed in companies is that of the Spanish fashion retailer group Inditex, owner of the Zara brand. Zara started in 1975 and grew into the world stage during the 90’s and 2000’s, which was the time that most established big fashion retailers had or were moving their supply chain to southeast Asia. The low labor costs in the region meant that the unit cost of producing apparels would be a small fraction as compared to the costs of local production even after considering logistics. During that time, the cost benefit was so obvious that the vast majority of fashion retailers went there. Armancio Ortega, founder of Zara, saw it differently. While the unit cost of offshoring was undeniably lower, he made sure to grow Zara’s supply chain as close as possible to the end market in Spain. This local production together with a well constructed information system to integrate the supply chain meant that while the unit cost would be significantly higher, it would also be faster!  In fact, while it took GAP or H&M a full 9 to 12 month to deliver a collection from design to stores, Zara reportedly was able to do the same in as little as 2 weeks. This agility is the competitive advantage that catapulted Zara as the top fashion retailer which in turn makes Ortega one of the richest people on the planet. This agility meant that Zara did not need to predict future trends, they could just react to customer actual demands, which in turn meant that Zara sold more stock without discounting or scrapping as it would not overproduce. Even to date Zara’s main competitors still struggle to match this agility. Zara does not have the lowest unit cost, but has the best system cost.

In the retail world, unit cost of products and services is the easiest metric to manage, so most focus is spent here. On the other hand, agility that provides system efficiency is driven at a strategic level and it is far harder to measure or project - this is why the best opportunities lie here. We need to refocus on speed to have the capacity to adapt to changes in customer behaviour and expectations and competitive landscape. To bring into a more practical example, today we see a few retailers moving towards a marketplace model in their ecommerces. This is surely a trend to keep in mind as offering products of external sellers augments the agility to increase product assortment, which improves customer experience and increases the monetization of the traffic. Today marketplaces are growing twice as fast as pure eCommerces as they can enhance assortment quickly, which results in faster growth, which allows them to service better.  

It is also true that when I talk to retailers about the concept of becoming a marketplace, they often think “Amazon” and even dismiss it for the perceived complexity and difference in model. However, the technology concept today allows for a more collaborative model, which is a “soft marketplace” where a retailer does not need to plan for thousands of sellers. Oftentimes, with one, two, or very few sellers, the benefits of the marketplace can be attained. One of my favorite cases is that of a Romanian fashion store Miniprix during the months of COVID lockdown. Within a few days of the initial confinement, they added a seller of basic food products which were products in high demand for their clients at the moment. With this move they increased traffic and managed to grow sales of clothing products as well due to higher interest and traffic. 

So the takeaway here is to look at the areas in your business where the historical focus on unit cost might have resulted in systematic waste - whether it is in the supply chain, human resources, processes and policies, or tools and technology. Which brings me to the last point… 

Technology

The final and critical pillar on the journey lies how well an organization can use technology to achieve business goals and agility.  

The reality is that retailers are in many cases ill-prepared to plan and purchase technology. Frequently, technology tools are jammed together in a very improvised way where the roadmap of implementations is a direct result of one person’s opinion (HIPPO above) instead of following a clear future architecture vision. Far too often, retailers are lured by buzzwords without a clear understanding of the technological concept behind it. To be able to build an architecture that drives diferentiation, it is critical to understand these tecnological concepts well and understand how exactly they can help to achieve your goal.

So here are some practical tips when it comes to building and agile technology stack:

Have clear goals

• Technology is a tool not a solution - you might hire the best tool, but if you cannot learn to use it or if you use it for the wrong purpose, it will end up bad. 
• Never move to select a technology partner if the goal to be accomplished is not clear. 
• Goals should clearly identify the sources of real differentiation.  

Assess the technical maturity of your organization:

Companies are likely to have failed projects when they are not aware of their actual capabilities or do not use them properly. Examples of problems are, an “early stage” company trying to select technology without assistance of an agency or consultant. Or an “Initiated” company trying to develop their own technical stack just because “the market leaders are doing that way”. Or an “Advanced” company that tries to develop in-house the entire architecture spending too much on areas that do not create differentiation and suffering with technical debt. The reality is that this lack of self assessment usually results in organizations making decisions as if they had world-class technology capabilities and failing miserably. 

Assess the industry technical maturity of each particular application

When defining an architecture and technology strategy, it is critical to assess not only the internal capabilities, but also the level of maturity of that particular segment. For example in eCommerce email marketing technologies, ERPs, CRMs, commerce platforms are highly mature technology areas. There are thousands of options in the market for all types of needs and niches. It is very likely that if you are developing or maintaining your own email marketing capabilities you are wasting precious resources. On the other side we have technologies of actual artificial intelligence, voice search, computer vision that are far less mature. There are early entrants and several innovative projects, but no established players in the market and not many reproducible success cases.  

The importance of such assessment is to define the go to market strategy for each area. So if your company wants to explore neural network machine learning applications, it is very likely that there will be a strong need to make personalized developments. On the other hand, nowadays world-class organizations often hire external technologies partners and platforms to build their technologies in areas that they know are not differentiating.  

In the world of eCommerce platforms for example, our state of the art is today on the SaaS platforms. Global technology analysts such as Gartner, and IDC already outrank the new generation of cloud native platforms as such platforms are already flexible enough to adapt to business needs but bring a plethora of benefits such as low total cost of ownership, faster time-to-revenue, and agility to the business. While there are still clear uses where an On-premise solution might still be the best, the majority of the cases today SaaS solutions are far superior in the current market maturity level. However, it is still the case today that in many markets, on-premise solutions are still dominant due to the status quo or the “unit price” thinking. It is easy to compare the obvious costs of license, and project costs which is the trap here… it is harder to project all the costs of evolution, agility and maintenance - aka the system costs - that make SaaS platforms generally better, and more agile architecture when SaaS platforms provide enough customization capabilities. There is a great example on how a regional groceries chain was able to grow by 10 times their online sales in 5 days during the confinement by leveraging the flexibility of a SaaS architecture. 

Never forget about Technical Debt

Everytime you create a custom application you create a debt. Such debts must be serviced for the life of that application. One of the top recommended books “The Lean Startup” by Eric Ries taught a new generation of software companies that one of the keys is to minimize features to the ones that truly bring value as each feature builds complexity that can. This is a great read towards this journey of digitalization. I have often seen companies engage in naive developments of open source projects or custom developments when it was not needed with the mindset that once developed it would be “free”. More often than not such projects fail or become a cash bleeding operation. The problem is that it is often easy to develop the initial delivery then there are the evolutions and changes to the project. Then the team changes. If you had 3 developers on the project and one leaves, suddenly 1?3 of the code is basically foreign to you. As time passes, there are more and more patches as early architecture is rarely appropriate for future requirements. Finally, the development reaches the inevitable end where it becomes too costly to make any improvements. When this point is reached without proper planning on a critical application of the company, the company slows down significantly and enters survival mode. It is critical to choose technologies that have demonstrated constant evolution capabilities and always budget for future evolution needs when considering a development. It is easy to find examples of old eCommerce leaders stuck with legacy technology - losing marketshare. Technical debt is one of the main reasons that powerful brands and operations will fail in the future so understand it and stay ahead of it. Don’t reinvent the wheel! Focus solely on the developments that bring differentiation. 

I hope that these tips will help your journey in the coming years and I welcome you to share this article and contact me if you would like to keep the discussion going. I would love to learn new cases and new points of view!