Experimentation / CRO

Experimentation and Testing Programs acknowledge that the future is uncertain. These programs focus on getting better data to product and marketing teams to make better decisions.

Research & Strategy

We believe that research is an integral part of experimentation. Our research projects aim to identify optimization opportunities by uncovering what really matters to your website users and customers.

Data and Analytics

90% of the analytics setups we’ve seen are critically flawed. Our data analytics audit services give you the confidence to make better decisions with data you can trust.

Ecommerce optimization series #7: Hit big hairy audacious goals by getting aligned

Dr. Edwin Locke and Dr. Gary Latham summarized 25 years of research on goal setting into the popular book, “A Theory of Goal Setting & Task Performance.” Their key findings were that “specific and challenging goals led to higher performance than easy goals, "do your best" goals or no goals.” The operative word in this quote is “challenging,” and likely why the concept of “BHAG,” which stands for Big Hairy Audacious Goals, was proposed in the 1994 book "Built to Last: Successful Habits of Visionary Companies" by Jim Collins and Jerry Porras.

BHAGs are very ambitious, compelling, and long-term (think ten+ years) but should still create urgency today. Examples include;

  • Best Buy’s BHAG is to move from an in-person shopping experience to becoming a digital leader in technology. 
  • Nike’s original BHAG was to crush Adidas.

Other goal-related research, such as that by Dr. Gail Matthews, found that people are 42% more likely to achieve goals if they are written down. 

[Side bar: A little known BHAG for Speero is that we want to be for experimentation what NNgroup is for UX. It’s the latest iteration of where we’re pointing the agency. We evangelize this internally, get the entire team educated on CXL’s courses for testing and CRO. It’s working. We’re more and more speaking the same language and the team is organically ‘filling in the gaps’ of our systems and frameworks. More to come on this soon.]

So specific and challenging written goals seem like a no-brainer, and I can’t think of many businesses that don’t have them. But how do such goals relate to the work of the experimentation team? Let’s dive into this topic to understand what’s going on.   

Why does ‘goal alignment’ need your attention?

Goals across departments must align with each other and the broader business; otherwise, Goal conflict undermines performance if it motivates incompatible action tendencies.” While your team might have a solid set of goals, other departments might be pulling in a different direction, canceling out your efforts. 

“One team is playing basketball on another team’s chessboard”

This is where a goal tree comes into play. A goal tree is a logical and hierarchical visual representation of goals, critical success factors, and necessary conditions. It’s a tool to understand goal interdependence and ensure everyone works towards the same overall purpose. 

 And it should remind you of an OKR setup. But an OKR system is more of the ‘GPS’ or navigation system, whereas the goal tree is more a representation of the speedometer metrics. 

Example Goal Tree 

But OKRs can be on the same tree. A goal tree can also break down your longer-term, super ambitious BHAG into a shorter timescale with more specific, quantifiable goals and tactical metrics per department. 

Ensuring your experimentation goals align with BHAGs, helps your team move beyond simple optimization work to driving significant strategic growth. It also changes leadership's impressions that experimentation is limited to slight tweaking that improves web metrics to a driver of substantial and strategic business change. For example, experimentation team alignment with BHAGs could help move the team goals;

  • From (unaligned); increasing conversion by improving UX and user flow on the website. 
  • To (aligned); increasing market share by experimenting with new technologies, positioning, and pricing that works for a new customer segment. 

The second reason goal alignment across departments is so important is that different calculations or data sets are often used within the same business. Research into ecommerce businesses found that departments tend to calculate profit differently, for example. 

And It’s not just profit calculations. Consider how your customer service department might measure customer satisfaction–perhaps calculated on a 5-star scale after a customer contacts them versus an ecommerce team who might use NPS or include online rating sites as a data point. How can a business reach its customer satisfaction goals in this scenario? 

So goal alignment is clearly an area for improvement for many ecommerce businesses. It’s also one area where we see high-performing (mature) experimentation programs leading the way (moving from tactical CRO work to business transformation through experimentation.) 

So let’s look into what goal alignment can do to improve the maturity of your experimentation program and the broader organization. 

The Opportunities

  1. A way to deal with conflicting priorities 

Let’s say your company's 2023 goal is to generate 5% more profit. It might feel like all departments are on the same page. But how do different teams approach this shared goal? 

The CRO team might decide their overall department goal is to convert more users to increase sales revenue and thus profit. One strategy they might employ is offering free returns to increase conversion rates. But how does this impact the operation department's goal? Their approach might be to reduce costs, one of the most considerable expenditures being returns. In this scenario, we face a conflict of priorities, and even though both teams are aligned on the overall company goal, their department-level goals, KPIs, and subsequent strategies aren’t. This is a great example where conducting a goal tree exercise across teams can promote cross-departmental collaboration rather than teams having conflicting priorities and working against one another. 

I’ve seen this misalignment of goals manifest in arguments over the website experience. Buyers want the homepage to feature their newest product line, which they believe will open up a new customer segment. Marketing intends to optimize its spend, so wants the homepage to reflect the campaign messaging performing well in paid ads. Welcome to a Frankenstein homepage that’s a cobbled-together mix of messages, products, and positioning, resulting in it not working well for anyone.   

Creating a goal tree in collaboration with departments across the company can help establish what everyone is working towards and prevent conflicting approaches. 

Another, more programmatic goal tree map example. This one is in Airtable, and you see it allows you to link metrics to what experiments were run to effect that metric. Thus automatic reporting.

  1. A focus on the metrics that contribute to growth 

The hierarchical structure of a goal tree means you are forced to link the KPIs/metrics needed to reach a specific overarching goal. Without presenting the relationship between goals and metrics, senior leaders might look at results from the experimentation team and struggle to make the connection. For example, what’s the relationship between conversion rate increases between stages of a digital journey and overall profit? It might be that this insignificant metric leads to an improvement in retention, but without a goal tree to show this, it’s hard to understand its importance and the larger picture.  

The nature of A/B testing also means we can’t always measure longer-term metrics we aim to impact within the test window, making it hard to establish whether our work is moving the growth levers of the business. The goal tree allows you to show the relationship between proxy metrics we might need to use to gauge our impact on the company's overall goals. 

Additionally, just because you can track and measure a million different data points about your digital experience doesn’t mean you should. But I’ve often heard, “If it’s no extra effort, let’s measure it.” This results in dashboards and reports that are overwhelming and underutilized because they don’t inform action or measure KPIs related to goals. 

What are the data points you need to measure the impact on your goals? Goal tree usage allows you to focus on the metrics that matter and quickly identify gaps in your measurement. Having such conversations also allows teams to uncover inconsistencies in calculating essential metrics. 

Experimentation is in the DNA of Uber Eats and a cornerstone of how we develop product.  We don’t need an experimentation center of excellence or ambassador prrogram because it’s just what’s done by every team.  But the challenge, according to Daniel Layfield, Product Manager, is thus too much data, nearly every experiment on the consumer apps will get a stat sig result on SOME metric.  Having a proper strategy, strong hypothesis and clear trade off frameworks is critical to not get stuck in endless post mortem conversations.

  1. A consistent strategy 

Have you worked in a business where your department goals or KPI change every other month? It’s far from ideal. 

While I’m not against adapting goals based on data and learnings, the changes are often knee-jerk or don’t give the strategy employed a chance, which means you’ll flip-flop between approaches.

Having clear, ambitious, and relatively stable goals means that as an optimization team, you can refine the discovery/research stage to focus on uncovering insights that relate to a specific goal. 

For example, Etsy has the BHAG “to make Etsy the starting point for your e-commerce journey.” 

Etsy’s company goal helps refine where and what to focus on as an experimentation team, and I’d presume their CRO department had shorter-term goals that work towards this company BHAG. E.g., the experimentation team’s goal might have been to improve product findability by 5%. This goal would then refine their focus, research, and subsequent strategy. This is perhaps why the Etsy team announced that they started using a technology called  “XWalk, to narrow the semantic gap by relying less on listing taxonomy and more on buyer interests…This means we can use 16x more real-time data to capture semantic, meaning across our inventory with XWalk than we could with our prior search capabilities.”

Clear department goals and KPIs also mean you can adapt your test prioritization model by weighing hypotheses that impact primary or secondary KPIs higher than those which don’t. 

You can orient your AB testing and product management strategy accordingly. In one of my favorite examples, Chetan Sharma of Eppo illustrated beautifully how incentives change with metrics,  


How does this all apply to your business? 

There are solid reasons and opportunities to align metric strategies, but the biggest is the decentralization and thus localization of decision making in organizations. You can align with leadership’s visions and goals all while unlocking faster decision making and action at all levels. 

This article is the seventh installment in our nine-part ecommerce optimization series highlighting the most significant opportunities and challenges facing your business. 

If you'd like to learn more about the maturity of your experimentation program and how you can assess and improve it, contact me at ben@speero.com

Look out for the 8th installment of the nine-part series next week.

Related Posts

Who's currently reading The Experimental Revolution?