Guide to growth-driven design

Large up-front cost

The average small to medium-sized business website typically costs anywhere between £15,000 - £50,000, a substantial up-front cost.

Whilst many web hosting sites, designers and developers will talk about much lower costs on their websites, these examples are usually very basic and unrealistic.

A fully costed, commercial website that meets business needs will be many times those costs. The true cost of developing a website is likely not an insignificant proportion of your marketing budget. And it is also paid in full before even knowing what impact the website will have on your business.

Large time and resource commitment

In addition to the up-front expense, the average SME website typically takes three months or more to complete and requires a great deal of resource and energy from your team. With no possible return on investment until after it launches.

Over budget, often late and inflexible

Even if the budget and time is approved, there are so many moving parts, people and steps involved in a website project that it’s difficult to estimate the cost accurately and to determine exactly how long the project will take.

Unfortunately, many website projects are delivered late and/or over budget. This would not only delay any results or feedback from your website, it would also reflect poorly on you in the eyes of your partners, whoever you report to, and other department heads.

Because new websites are usually delivered to a rigid brief without the benefits of either past experience or feedback, they are often badly matched to the real world and difficult to realign once launched, due to the inflexible nature of their structure.

No guarantee of improved performance

You will be held accountable for a measurable increase in results from your new website. So then the question becomes: after all the time, money and resources you’re putting into your website redesign, how do you (or the agency you’ve hired) know that what you’re finally launching is going to be the best possible performing website?

The answer is you can’t, it’s impossible.

All you can do with traditional website redesign is look through any past usage data you might have, conduct, or commission user research and then formulate a hypothesis of what you believe will be a high-performing website.

Then your hypothesis is implemented and never validated to see whether your thinking was indeed correct. If it’s not, well, we’ve all heard horror stories of websites being launched and performance tanking for one reason or another.

Websites typically sit with no major updates for 1.5 to 2 years

Whatever the reason is; “No Time”, “Spent all our budget”, “Other Focuses”, etc; we often let our website, our #1 marketing asset and best salesperson, sit relatively unchanged for years.

The problem is, that if too much of your site remains static, then your SEO will suffer and visitors will lose their trust in you. This is clearly not an ideal way to maximise website performance, yet it is common practice. Yes, you may make some small updates or improvements, along with adding blogs or landing pages to the site, but the vast majority of the site probably remains untouched.

The shifts marketers need to make

It’s time we take a step back and look at how we can approach the website redesign process from a different angle.

We need to find a superior process that avoids all of the risks inherent in the traditional web design process. We need a new process that produces a peak performing website; A web design process that is quick, agile and produces better results and ROI.

That process is growth-driven design.

Much like in a traditional website design process, the first stage is strategy.

In this stage, we’ll develop a solid foundation that we can build our process upon using the following steps:

Strategy

Goals: What are the performance goals that we are trying to achieve? How have we performed historically, where would we like to improve, and how will this impact marketing goals?

Personas: Next, unless we have already done this during your Inbound Strategy Blueprint, we will develop detailed persona profiles for the different types of groups visiting the site. A persona is a fictional representation of your ideal customer. You can create different groups of personas based on common characteristics your audience shares. This could be a point of pain, industry, job title, etc. Growth-driven design centres around the user, so it is critically important to fully research and develop your persona profiles in the beginning, as they will set the stage for all future activities.

Website and analytics audit: It’s time to start digging into the data. Perform a quantitative audit of how the existing website is performing, reviewing what is, and is not, performing well, where users are dropping off, etc. As we are completing your website audit, we will start identifying where there is opportunity for improvements for future web work.

"Our users are constantly telling us what they like and don’t like about our websites and apps through the actions they take. Quantitative research is when we listen to our users by collecting data for these actions and interpret in aggregate what the data means about our user base or product.

Through interpreting our quantitative data, we can create a hypothesis about what we can improve, how we can improve it, and relatively how much impact our improvement can have. We can then test our hypothesis with an experiment and use our quantitative data to measure the results."  Matthew Rheault - Sales Pro, Growth Team - Lead Developer

User research: After we have identified some of the areas of opportunity though your audit, we proactively reach out to your existing users to learn more about them, gain a better understanding of who they are and find ways to improve. As we’re collecting new user research, it will help us validate the assumptions we put in your original persona profiles and will probably give us additional information to include.

“Through qualitative research, we’re able to observe the user’s goals, motivations, and pain points in action. This helps us to develop an understanding for the human behaviours that are tied to the quantitative data points that we’ve collected.

As a result, we can begin to empathise and design for the human on the other side of the monitor, thus improving our user experience and moving the needle on key performance indicators like conversion and retention.” Austin Knight - HubSpot - UX Designer

Fundamental Assumptions: Using what we’ve learned in these previous steps, we start forming some fundamental assumptions about your users.

Some examples of fundamental assumptions might include:

• Value propositions for each product, service and offer
• The various locations and devices users will be accessing your website from
• What information your users are looking for

These fundamental assumptions will help us explain the behaviour and motivations of your users and will be influential in both the global and page strategy and future growth-driven design cycles.

Global & Page Strategy: The last step in the strategy phase is to develop both a global strategy for the website as a whole and a specific page-by-page strategy for each major page on the site.

Both the global and individual page strategies will derive from all of the previous steps and lay out a detailed strategy of exactly how best to engage and influence visitors to attain your goals.

Wish list

The next stage is developing your wish list. Taking what we’ve learned in your strategy planning, we brainstorm every impactful, creative and innovative idea that you’d like to include on the site.

The key is to come into the session with a “clean slate” and to avoid getting hung up on the existing website. Think about what items should be on the list to achieve your goals in an ideal world if money, time and development skill were not an issue.

This includes brainstorming ideas, such as:

• Impactful website sections and pages
• Marketing assets, tools and resources
• Specific features, modules and functionality
• Design elements
• Changes in experience based on devices, country, etc.

After a few hours of brainstorming, we should have a wish list of 50-150+ ideas for the new website.

This wish list will not only be used both to determine the initial action items to implement on the new site but will also work as an agile and flexible list that we will continuously add to (and subtract from) as you re-prioritise actions items over time.

Launch pad website

In the traditional web design process, we think of launching a website as the finish. With growth-driven design, it is just the beginning.

At this stage, we will be building and launching what we call a ‘Launch Pad’ website. This is the starting point from which all of your other growth-driven design activities and improvements start from.

The Launch Pad site should be launched quickly and will not be definitive. We want to avoid getting stuck on analysis, features, or content while building our launch pad. It may not be perfect on launch and it doesn’t need to be. But it will probably be a big improvement over your current website and creates a solid foundation from which we can continuously improve.

The size and complexity of the Launch Pad website will vary depending on what you have on your wish list and what type of website you need. However, it’s extremely important that we’re able to boil it down to the essential twenty percent that will make an impact and launch quickly so we can continue to learn about your users and improve the site.

Run 80/20 analysis on your wish list

In the wish list phase, we compiled a long list of all the action items we’d ideally want on the site. Now it is time to start sorting and prioritising these wish list items to determine which action items are the first ones to implement.

We will review the list and identify the twenty percent of items that will produce eighty percent of the impact and value for your website’s users. Once we have identified those core twenty percent of items, we pull them to one side and do some additional filtering by asking yourself, is this action item…

• A ‘must have’ or actually a ‘nice to have’? – If you answer ‘nice to have’ then it will return, back to the main list.

Then, with the remaining items, ask:

• Is this absolutely necessary for the initial Launch Pad site, or could we build it into the site in month three or month four?

The goal of asking these additional questions is to narrow your focus to the core, ‘Must Have’ action items that will provide the most impact. It’s essential to narrow down to these core action items to ensure a quick launch.

Hypothesis statements for each core action item

Once we’ve narrowed our list of action items for the Launch Pad down to the core twenty most impactful, ‘must have’ items, we will then create a ‘hypothesis statement’ for each one of the action items.

The hypothesis statement allows us to gain clarity on how each action item relates back to the goals we’re trying to achieve, the persona we’re focusing on and the expected impact this change, or update, will have.

Action item: Hypothesis statement

For [Marketing Matilda] visiting the [Pricing Page], we believe changing [Platinum Pricing] to [Request a Quote] will [boost lead conversion by 10%]

We believe this to be true because [research or previously validated assumption]

Expected Impact + Effort Required + Metrics Measured + Definition of Complete

At the bottom of each statement, there are four important items:

• Expected Impact: The impact value should be a single number based on the value the visitor will get from the action item and the impact that will have in moving toward your goals.
• Effort Required: The effort required should also be a single number that represents the combination of the number of hours, resources and difficulty to implement that particular action item.
• Metrics Measured: What metrics will we need to measure to test this action item and evaluate if our hypothesis was correct? The more specific the metrics, the better.
• Definition of Complete: What are all the steps we need to complete in order to consider this action item complete? Defining this up front is important because it will erase any grey areas that may arise down the road when reviewing results or measuring efficiency.

Continuous Improvement: Once we have launched your Launch Pad website, it is time to commence ongoing cycles of continuous improvement. Experimenting, learning and then improving the website. Coming out of the Launch Pad phase, we will still have a wish list of impactful items that we’d like to implement on the site.

This list should be regarded as agile and should therefore be updated on a regular basis.

It all revolves around your personas: A cycle starts with and revolves around, the personas who are the target audience for your website. At each stage of the cycle, we must continuously ask ourselves how each element relates and provides value to these personas visiting your website.

During this process, if it ever becomes unclear how an action item provides value to, or relates to, the persona, then we must take a step back and re-evaluate our thinking and even question the inclusion of that action item.

Cycle Step 1: Plan

The first step of the growth-driven design cycle is planning. Here we will be identifying the most impactful items at the current moment and planning to implement the most promising ones into the current cycle. There are a number of steps to go through in this planning phase:

Performance vs. Goals: Review the current performance of the website and contrast that to the goals we’re trying to achieve. This will inform us of any opportunities for improvement.

Additional Data or Research: Coming out of the last cycle and while reviewing performance vs. goals, there may be additional data and research we need to undertake to clarify new action items that should be added to your wish list.

Learning from Marketing & Sales: This is the point at which to consult with the marketing and sales teams and see what key items they have learned about the user since the last cycle. This information can hold valuable insights that can then shape the action items we’re implementing in your growth-driven design programme.

The marketing team may have written a blog on a particular topic that exploded in popularity and resulted in a great number of organic and social visitors. We’ve now learned that this topic is important to your personas. How can we take that knowledge and add new items to your wish list to influence your growth-driven design programme?

Brainstorm and Prioritise Wish list: Based on the new data, research and learning up to this point, we will facilitate another brainstorming session to determine any new action items to add to the wish list.

Your action items will fit within these areas:

Boost conversions: The first group of wish list activities are those that are directly related to conversion rate optimisation.

Consider:

• Conversion points
• User path
• Testing value props
• Split testing

Improve user experience (UX): Improvements to the website that give the user a better experience and make it easier for them to navigate, find what they are looking for and solve their problem(s).

Consider:

• Navigation
• UI improvements
• Mobile experience

Persona type (personalise to the user): Adapting the site, calls-to-action, content offers, etc. to the specific visitor based on the data we know about them. This includes, but is not limited to, tailoring based on interests, type of contact, device, geolocation, referral source and previous actions on your site.

Consider:

• Calls to action
• Previous behaviour
• Lifecycle stage

Build marketing assets: Marketing assets are items that hold great value for your marketing programme, such as email lists, social accounts or your blog list. You can build new marketing assets into the website such as tools, in-depth resource sections, online training, or directories - any item that will provide great value to both the end user and your company. A great example of a marketing asset is HubSpot’s Marketing Grader. Users get a detailed report of their digital marketing efforts and HubSpot gains links, data and leads.

General website updates: There will also need to be more general website updates that come up from time to time that can be added to your wish list as well. These might be product or service updates, company restructuring, and so on.

Prioritise your wish list: Once we have all the new items added to the wish list, we will again prioritise all the action items based on the (High / Medium / Low) impact they will have on the goals of the website and value to the user.

Plan sprint cycle: With an updated and prioritised wish list, we can pick the most impactful action items that you want to implement in this cycle. The number of items we pick will depend on how long the cycle is. It is better to pick fewer items and focus on implementing those items effectively. If these are completed ahead of time, then we can always go back to your wish list and pick more.

Cycle Step 2: Develop 

Moving to the ‘develop’ phase of the cycle, we now have the most impactful action items to work on and it’s time to start implementing them on the site. This is where the rubber hits the road and we start completing each action item that you selected in the planning phase.

Each action item that we implement should be considered an experiment to see the impact it has on the performance of the website. To measure these experiments, we must setup validation tracking around the metrics outlined on the action item.

After the experiment is pushed live, we may want to develop a marketing campaign (social, PPC, blogging, etc.) specifically to drive traffic to that section of the site so we can start collecting data. If that is the case, then during the development phase of the cycle, we will build and schedule that marketing campaign.

Cycle Step 3: Learn

After the experiments have had enough time to run and collect data, we move on to the learning phase. Here we are going to review the information collected on your website visitors.

Based on this information, we then validate or disprove the hypothesis associated with the action item. Did the change have the impact we expected? If not, then what could be the reason? Based on the results, what did this teach us about your visitors? What did we learn that we didn’t know before?

Once we determine whether our hypothesis was correct and laid out what we’ve learned about users, we’ll want to publish this information in a central location so everyone within the organisation can benefit from the insights. Having a structured system for publishing your findings is also a great reference for the future if you ever need to look for trends or reference previous experiments.

Every website or application has a unique subset of users that share common traits. It is a primary goal of growth-driven experiments to answer questions about our unique subset of users.

Let’s imagine that we have a food blog and want to know how to best engage our readers. Take a look at the following example learning we’ve gained through running quantitative experiments:

• We learnt that our users want comfort food in the winter, but want to lose weight through the rest of the year.
• We learnt that our users are more likely to read an entire piece of content if our first paragraph contains less than 120 characters and is preceded by a full width image.
• We learnt that our users are most engaged with our content at 9AM on weekdays.

"These lessons all compound on one another and help us to hone in on who our users are and exactly what works best to engage them. With these three key learnings, we will know what to write about seasonally, how to format our content, and when to send out our newsletter." Matthew Rheault - Sales Pro, Growth Team - Lead Developer

Cycle Step 4: Transfer

The last step in the cycle is to transfer any impactful information we’ve learned through the process to other parts of your business.

We take time to review what we’ve learned from each completed action item and brainstorm how this may be useful for others. And we review previously completed action items to see if we can find any patterns across your users.

Here’s an example:

Let’s say that in your experiment, we were testing two different landing page variations. One variation used social proof as the independent variable and the second variation used authority. After running the experiment, we reviewed the data to find out that for users coming to your website, social proof was a much more influential factor leading to conversions. Now that we’ve learned that social proof is a better trigger for your audience, you can inform your other teams to start incorporating social proof into other parts of their marketing and sales activities, such as email or sales scripts.

Once we’ve put together recommendations for others within the organisation, you may want to host a meeting to update them with the latest website insights and then brainstorm ways that they could integrate and transfer this data into tactical action items – where relevant – within their own departments.

Then, Repeat!

Once we’ve completed a cycle with a set of action items, we go back to the beginning and start planning your next cycle. And so the cycle repeats itself over and over again through a process of continuous improvement across each of the steps. Every iteration improves the effectiveness of the site and provides additional knowledge about your visitors. The more cycles we can complete, the more focussed and effective your website will be and the more impact your website will have on your company’s success.