Episode #3

[Webinar] Conversion Optimization is an Operating System

Speaker: Alex Birkett
CO-Founder of Omniscient & SR. manager, Experimentation at Workato

Conversion Optimization is an Operating System (Not a Tactic): Alex Birkett

Alex Birkett is a leading marketer and experimentation leader at Workato. He has his own premium agency named Omniscient Digital. He has worked with companies like ConversionXL and HubSpot in the past. His expertise and experience lies in content marketing. He loves to compose and listen to music and lives with his dog, Biscuit in Austin.

Highlights of the Webinar:

  • Discussing CRO as Tactics or Strategies
  • Why CRO can be better understood and implemented as an OPERATING SYSTEM**
  • Centralized and Decentralized Models**
  • The two components of CRO as an operating system- technical and attitude/cultural**
  • Decision Making on the basis of risks and certainty**

Today, I’ll be discussing how conversion optimization is an operating system, not a tactic or a list of tactics. However theoretical it may sound, its practical implementation will definitely boost your conversions and the success of your marketing efforts.

Is Conversion Optimization a list of Tactics or a Strategy or Neither?

When you try to find out about conversion optimization, there are lots of articles that list tactics and tips to improve your conversions through various channels and campaigns. Discussing the tips on this list, we see that these are regarding how one can apply videos and pictures or even marketing copy so that we get the maximum benefit. However, these are debatable aspects, for example- to what extent videos will succeed for B2B or e-commerce. On these lines, I feel that the implementation of these suggestions is limited. Also, in the e-commerce industry, there can be variations in how one can apply these tips and the corresponding success.

Let's discuss how and why we should perceive conversion optimization as a strategy. There is a thin line between what is touted as strategy and tips for example- FOMO and social proof. And once again, these have multiple limitations depending on the product and other resources. Also, the outcomes of these are not as far-reaching as they should be.

I have written tons of blogs on conversion optimization. My knowledge and expertise come from my work and from my engagement with communities of conversion optimizers. I believe that a lot of people misunderstand conversion optimization as a tactic or strategy. I, on the other hand, believe that it is neither but an operating system.

Defining CRO as an Operating System

Strategy is the overarching set of goals or ideas and tactics are the specific things one can do to achieve the outcomes. Now, let’s look at what an operating system is in technical terms. By definition, it is the software that supports a computer’s basic functions, such as scheduling tasks, executing applications, and controlling peripherals. In the context of CRO, I am extending that to enable experimentation, quality control, and data-driven decision-making in a company.

This definition is best implemented when we visualize this in the form of infrastructure, with well-planned pipes and channels instead of thinking of this as a superhero vigilante who will take ownership of the entire process. It is the underlying infrastructure that facilitates the flow of knowledge and data.

This idea is inspired by an article on growth in which they say that growth is an operation and the underlying process. The writer says, growth is not just marketing, it is enabling marketing and also beyond marketing. This is how I redefine CRO-

“A method used to make better business decisions in an online environment and deliver better experiences to visitors, customers, users, etc.”

I look at it this way- there is tons of information and research data coming in from CROs and these can be used for decision making and thereafter application which is nothing but experience delivery.

Centralized model and Decentralized models for CRO

The centralized model is more common with the leadership at the center. In the decentralized one, there are different teams like email, social media that work on their own experiments and accomplish their goals. The latter is common in SaaS companies. Both these models have their advantages and drawbacks.

That brings me to my favorite model which is the Centre of Excellence model that is about Delivery, Technology, Process, Analytics, People. It aims at quality assurance and enablement. I believe this model is prevalent in mature organizations that prioritize experimentation and excellence. It facilitates the running of experiments, you can set up analytics and infrastructure, you own the process, it is more homogenous and predictable over time.

Two components of CRO as an operating system- technical tools and attitudes/beliefs (culture)

Component 1- Technical Functionality (tools)

  • Data operations- data collection, integration, accessibility, trustworthiness.
    Tools- Google Analytics, Amplitude, Wynter, Qualaroo, Segment, Workato, Snowflake.
  • Testing the trustworthiness of data
    It’s very important to check data, test, and question its poke holes and check whether data is telling what we think it is telling.
    Tools- AB test by CXL, Convert, Google Optimize, and Customfit
  • Knowledge sharing and program management
    It is often quite underrated, however, it’s crucial especially in the decentralized model wherein knowledge is supposed to proliferate to other teams.
    Tools- Airtable, Confluence, Effective Experiments.

Component 2- Attitudes and Beliefs

  • Evangelizing- Not everyone in the team is acquiring knowledge and skill-sets are also often different, however, sharing of responsibility is important. And so it's important to get people excited about the processes. Books on this subject- The Innovator’s Hypothesis, Trustworthy Online Controlled Experiments and Experimentation Matters.

How I do it-
a. Regular experiment review meetings
b. Email newsletter Reviews
c. Slack channels and Automated Notifications
d. Forms and Bots and pulling in other ideas (inviting ideas from other teams)
e. Celebrating Wins, Case studies
f. Gamification/ Competition (For example: incentivization)

  • Education-
    Not everyone will have the background of data skills and so it’s important to write wikis, documentation, conferences, and encourage your team also to do so and also to take up courses. For example, at Airbnb, they formed a data university. These can be the foundations for qualitative and quantitative data analysis.
  • Support-
    CRO is intimately connected to other teams like paid acquisition, SEO user activation, UX Design, Data and Insights, Email Marketing, Monetization, mainly in a decentralized model.

Three types of queries-

  • Navigational query- searching for the name of a website or way to go to a specific page
  • Informational query- searching for information on a given topic or intent, to learn about something
  • Transactional query- searching for a product or service to purchase. Intent to buy.
    These queries affect conversion optimization. A conversion optimizer must understand the intent of the visitor to know- 1) What action must be most appropriate, 2) What action is most likely to elicit a conversion 3) The best method of persuading the user to convert.

CRO and Paid Acquisition

These two teams are mostly intertwined. From the paid acquisitions, you know which keywords are working for you. You can study the keywords and create pages on the relevant topics. Run feedback surveys and polls to understand what people’s pain points are and what type of content you can create to help them and gain visibility among them. You can use the insights from paid activities to create organic content that will fetch revenue in the long run.

CRO and Email Marketing

Let’s say you were checking a product but didn't buy, so then, you get an email following up on your activity. It is quite effective to follow up through such kind of email marketing.

CRO insights help in decision making

We are in the game of reducing uncertainty and improving business decisions. There is no decision that is completely certain. Varying amounts of risks and uncertainties are involved. You have to weigh the possibilities and make smart decisions.

Summing up

All the planning and the corresponding efforts should aim to move the bottleneck, unleash innovation and facilitate data flow.