Forecasting is fundamental to everything we do

James Finlayson is Head of SEO at the UK’s largest independent media agency, the7stars, where he applies a mix of technical and creative skills in order to put the right content, at the right time, in front of the right consumer.

How do you usually approach pitching to eCommerce clients?

Any pitch starts by really understanding the challenge the company’s trying to solve, what they’ve already tried, and their brand voice and positioning. We believe it’s important to spend time assessing the strategic problem that unlocks the brief rather than rushing to an answer before insights have been sufficiently understood. 

Is the challenge getting on consumers’ radar? Do we need to drive those pre-disposed to the brand back on-site? Are there conversion opportunities? Are there negative perceptions around the brand that need to be assessed?

Then we begin to validate the same—building up key personas and industry-wide trends. Using a combination of social listening and keyword research, this turns into insights on product and category-level opportunities, mapped and sized to different audience states. 

Finally, these are collated into tactical short-bursts of activity to impact and supercharge key audiences and activities. These can be very varied, but two examples would be:

We recently worked on a project based around the key insight that getting existing customers to go in-store more often, rather than just staying online, led to better outcomes. As such we edited website content targeting this audience to mention the option of visiting in-store, whilst also improving the branch pages themselves so that finding the branches—and understanding all the services they provide – was easier.

For a beauty brand, we realized that the use of Google Images was way up YoY and that, in their niche, Pinterest was frequently appearing. As such, we put together a quick plan to create some Pinterest pages targeting specific ‘looks’ that we knew fit our audience and targeted terms that Pinterest was already ranking for in Google Images. It wasn’t a huge task, but it inserted the brand into consideration sets earlier on in the funnel, drove significant views and, as a result, potentially helped to drive purchase decisions.

Then, there is the long-term overarching strategic plan that’ll be the story of our work and will be where C-level buy-in occurs.

Throughout, we look at how different channels can layer their activities to reinforce and multiply each other’s ROI. At the7stars we’re lucky, here, to have unique insights to help with those decisions through proprietary tools, including our own consumer sentiment tracker and Gravity Planner, our award-winning multi-channel marketing investment planner built in partnership with Nielsen.

Do you use forecasting for pitching or upselling to eCommerce clients? How do you approach it?

Forecasting is fundamental to everything we do; never invest in anything without knowing the level of return and the risk associated with it.

Fully understanding incrementality and attribution has become more important over the years. Otherwise, forecasts over-sell and under-deliver. 

the7stars is the first agency I’ve worked at to have a dedicated Analytics & Insight team. They’re able to get involved with long-term cross-channel forecasting, run econometric studies, and surveying to understand client motivations and pain points.

Through this, we’re able to run multi-year forecasting that takes into account changing consumer behavior and concentrates on a destination we can all rally around rather than the more traditional, though less important, the question of ‘what can be achieved in the next 6 months.

The Analytics & Insight team uses a wide variety of tools, half of which were developed in-house, to run full econometric modeling. This includes datasets such as Nielsen and SimilarWeb, a fair amount of data science, and clients’ datasets from across the marketing stack. This is for big, complex forecasts, though. SEOmonitor’s forecasting tool is great for quick, indicative planning.

Sitting in the middle is a more manual process that takes into account the specifics of our planned growth; when are new pages likely to go live and how long is it likely to be before they start receiving traffic, what effects do we expect other channels to have on branded search, what is the deteriorating economy likely to do to conversion rates, are we likely to see product price increases? With this middle-ground forecast, we’re looking to think, month on month, about what levers will be positively and negatively affecting traffic and estimate accordingly.

What are the main SEO challenges your eCommerce clients are facing?

Changing consumer behavior. With the cost-of-living crisis and the changes a post-COVID society brings, historical search volumes have become poor predictors of future consumer behavior. That’s made budget allocation, growth, and focus areas both more difficult and complicated.

The COVID pandemic was a (hopefully) once-in-a-lifetime event of such magnitude that it fundamentally changed society. Millions made the move from city living to the countryside, over a third of workers now work from home and many simply retired early.

As a result, consumer behavior radically changed, and now, with a cost of living crisis, it is changing yet further. That means that YoY search volumes and consumer behaviors are likely to bare little relevance to the same one year in the future. If you don’t know your market size over the next year, it’s incredibly difficult to know how much to invest in marketing; or whether to invest at all.

Overcoming the above has led us to:

1. move towards providing a range in our forecasting rather than a simple estimate
2. provide more details as to all the assumptions that go into our forecasts
3. zoom out in our reporting – talking more about macro industry-level trends and how to react, and less on just what’s happening within the client’s site.
4. as part of the above, we’re investing in tools (and building out tech) to monitor competitor site changes so that we can learn from, and react to, competitor changes as quickly as possible

We’re also making it clearer that our proposed deliverables are more flexible than ever. The client’s paying for results, not documents, and our work may change from what we originally anticipated. That makes pivoting far easier.

the7stars also conducts a series of regular surveys on consumer views and behaviors. As a team, we discuss this new data as soon as its launched, thinking forward in terms of how any identified trends may affect our clients. That data gives us a license to recommend changes to clients that show our agility and tenacity to find a way back to the forecast no matter what happens.

A reduction in Google’s dominance. We’ve talked for several years about how Amazon appears in 10% of Google Searches, but now an increasing number of searches are beginning on Amazon rather than Google.

Younger demographics are now searching on TikTok more than Google. As a result, we’re increasingly being asked about Search Experience, rather than SEO – how eCommerce clients can dominate across search channels rather than ‘just’ Google.

Lack of resources. CMSs are becoming more complicated, and so any change requiring developer resources must justify its place in a sprint against other tickets. Meanwhile, brands find they have too few copywriters stretched across too many channels.

We’re tackling it by increasing use of a/b testing frameworks for recommendations; small tests to prove efficacy before asking for resources, allowing us to better estimate the value of the change.

What is a recent more specific challenge that you had to overcome in an SEO campaign for eCommerce?

One of the UK’s largest takeaway brands is currently going through a complex migration. The design philosophy for the new site put a single, simple customer journey at the heart of all their plans. Yet it relied on customers beginning their session on the homepage.

Using a combination of data sources, we were able to show the true multi-session customer journey across Google, the site, and other sites. That got us a seat at the table, but it was our keyword segmentation that allowed us to prioritize particular journeys and so lead collaborative efforts on evolving the design philosophy.

As a result, key SEO landing pages were added back into the plan and, our forecast shows, over £800k a year of income was protected.

Can you give me an example of a successful SEO campaign for an eCommerce client? Which were the main tactics and challenges? 

During lockdown, furniture companies saw large increases in sales, but with the country opening back up and hardening financial times, Sofology was keenly focused on where future growth would come from. 

So, our focus was clear (1) increasing traffic on directly converting terms – no fluff; and (2) working across channels to improve data quality, share learnings, and multiply the effects of our work.

We began by setting up unequivocal dashboards to keep everyone concentrated on the metrics that mattered. Supported by refreshed content across the site and brand-new buying guides, we increased the pool of commercial terms the site could rank for. 

Deeply embedding ourselves within the client’s development sprints, we were able to push through more than 2-dozen technical fixes over the course of two months that allowed Google to give full value to these changes. Finally, working with our PPC team, we were able to make micro-optimisations to increase both rankings and conversion rates.

In 7 months, the team:

• Achieved number one position for ‘sofas’
• Increased the percentage of rankings in the top 3 from 15% of our key terms to 88%+. That’s 1.4k keywords moving into the top 3 positions.
• Increased page 1 rankings by 88.3%, propelling 1.6k keywords onto page 1.
• Increased organic revenue by 35% during, and despite a dip in search volumes – a 26x ROI.

What have you noticed in terms of search behaviour trends within your 2022 campaigns? 

Google’s search volumes aren’t just stagnating – in many cases they’re reducing. Despite this, consumers are conducting more research prior to purchases than ever. Those searches are happening on Amazon and other ‘super-retailers’, on TikTok, large publishers with loyal audiences, Pinterest, YouTube, and Reddit. 

We recently looked at one market where, we estimated, less than 10% of search activity was actually happening on Google. 

Except for seasonal bumps, searches for ‘sofas’ have been dropping since April 2021 (currently down 13% compared to Nov 2020), searches for ‘milk delivery’ seem to have reduced every day since we left lockdown, searches for ‘pizza delivery’ are down 54% in the last 2 years, searches for ‘theatre tickets’ are down 10% on pre-COVID levels and searches around dating, as a topic, are down nearly 20% compared to pre-COVID levels.

Have eCommerce clients’ needs or requests changed over the past months, given the economic shifts? If yes, how? 

Everyone’s under a lot more pressure to perform and prove ROI on their work. That, inevitably, leads to a requirement for more and clearer forecasting, an appetite for safer strategies, and increased questions/requirements for channels to work collaboratively.

That same risk aversion has led to development projects involving SEO earlier in the process. And so we’ve been able to bring more value to these projects by shaping their growth, baking best SEO practices rather than trying to fix them shortly before (or after!) launch.

How do you approach SEO budgets in eCommerce in the context of 2023? 

Multiple econometric studies, and recent case studies from COVID spending changes, have proven that in a tough market, when competitors are tightening their advertising budgets, outspending competitors almost always provides both short and long-term benefits across sales and awareness. That’s often counter to C-level instincts, though. As a result, we’re focussing on:

1. Increasing the transparency of SEO’s incremental value; both in terms of data-driven attribution of conversions and equivalent advertising value.
2. Pushing for larger, not smaller budgets. SEO tends to scale well, with increasing efficiencies as budgets increase. As such, larger SEO budgets often provide a much higher ROI.
3. Solving business challenges that happen to be best implemented through SEO. That gives SEO Managers and CMOs options as to what budgets they can access to support the activity.

We recently advised on a project where a client has a big paid push in April next year. We found an opportunity to build out some on-site informational content that would help grow their re-targeting audience buckets ready for that paid spend.

In another project, we helped a client’s merchandising team with work on what they were planning to put on the back of their revamped product box – based on keyword research, persona creation, and an understanding of key consumer pain points.

In another project, we advised on the purchase of a client’s competitor. For that project, the insights were around market size, market capitalization (across channels) and, mainly, risk.

We showed how and where the industry was growing/shrinking, how new competitors were growing quickly, how sentiment around the brand wasn’t particularly positive and how their previous marketing activity created ongoing risk.

All of the above relied on core SEO skills, and everyone in the team felt very comfortable working on the projects – but they weren’t directly SEO projects. They were the most effective things we could do for the client at the time – and they allowed us to access non-SEO budgets.

Have you noticed any differences between different industries under the eCommerce sector, in terms of search behavior trends? 

Luxury-goods retailers are typically finding a more challenging market than those who’ve traditionally been a necessity. 

Equally, industries that are price-sensitive have seen increasing traffic but reduced conversion rates as people are shopping around more.

Fashion and jewelry searches are becoming ever more visual in their search behavior, seeing an increased trend towards Google Images, Pinterest, and TikTok.

How do you leverage SEOmonitor for eCommerce clients? Give me a few examples 

SEOmonitor is a key tool for us. 

The smart grouping functionality allows us to split the keywords in a myriad of ways; by SERP feature, conversion rate, site section etc. That makes it much easier to see macro trends that might be causing keyword-level changes.  

Finally, the pitching account functionality provides really valuable flexibility. As a team, we’re constantly throwing around ideas for client plans course-corrections. Being able to quickly throw a few thousand keywords into a project to model out the opportunity, at no extra cost, encourages data-based thinking and all the right kind of experimentation.

The forecast product is useful to quickly:

1) look at the feasibility of traffic growth numbers in RFPs – are they even possible with a ‘standard’ strategy. What more can alternatives offer?

2) prioritize work. If we’ve four things that we’re recommending, we’ll create a folder for keywords related to each and then see what sort of range of growth each might be able to provide. Is one strategy high risk (in terms of implementation and other factors), but high reward (in terms of traffic generation)? Would covering off two of the smaller tactics produce a higher ROI?

For the first point, the purpose is to simply understand the ballpark feasibility. The second point is the comparison rather than the hard figures that it provides. Social listening comes into play when we’re being asked to provide absolute figures that the client can then use in their own marketing plans; in that instance, specificity is far more important, and so we spend the extra time to estimate future trends and topic growth using social listening.

___