E-commerce

How an e-commerce benefitted from Adsellum's system to manage its well-curated advertising campaigns, increasing sales and improving financial ratios.

Initial problems

This new e-commerce website faced many common problems of new online shops, as well as a few specific ones:

  • The brands offered on the website were not well-known and do not generate enough organic traffic to drive sales. Competitors’ brands were using the same marketing elements to sell.

  • SEO efforts may have taken too long to implement and produce results for the main product lines, given the need to sell quickly and in large quantities. The market was already saturated with established sellers with better-stocked and better-ranked websites.

  • A majority of secondary products will eventually be properly referenced, but these product lines are priced too high relative to the competition and do not generate sales for the main products. Instead, the main products tend to lead to secondary product sales.

  • A particular brand name was negatively impacting SEO, penalizing the entire website.

  • The website’s main products are on average 20% more expensive than the competition, and sometimes much more. The products were of higher quality, but this argument did not carry as much weight in distance selling. During a recent promotion period with a 20% discount, conversion rates returned to normal for the sector, confirming that the pricing was a problem.

  • Demand was increasingly focused on the website’s specialized products which happened not to be the most profitable to sell.

  • Tracking was difficult to perform since a large part of the orders were carried out by telephone or email exchanges rather than through the online shop funnel. Analytics solutions were not collecting enough sales data anyways.

Rapid customer acquisition

Online advertising

Decision was made to advertise on Google Ads and Bing Ads (by automatic import). It was necessary to focus on transactional rather than relational marketing/advertising. Later on, relational advertising was found to be ineffective anyways.

Given the difficulty in tracking actual order amounts, the decision was made to optimize CPA rather than ROAS. Significant actions, such as form submissions, were tracked reliably. Moreover, telephone calls were also accurately tracked using a redirection phone number on the main advertising platform.

SEA and Shopping campaigns were implemented. These primarily targeted highly qualified prospects before being expanded once managed by Adsellum’s system.

The SEA campaign had an excellent CTR of 20%, meaning that one in five people clicked on the ads. This was due to well-written and precisely targeted ads. By comparison, the sector average was 3%. However, sales remained low due to the high product prices, which strongly impacted conversion rates. Despite very high quality scores and other good metrics, the CPC was also often high, due to the highly competitive sector.

Initially, the main SEA campaign generated sales with a CPA of €50 and an average order value of €105, resulting in a ROAS of 2.1. These values fluctuate seasonally, with demand three times higher in winter than in summer. Even with a low conversion rate, the ROAS should have been higher given the degree of targeting precision. However, the rest of the website’s problems were negatively impacting this metric.

Furthermore, the Shopping campaign’s performance was impacted by excessively high call prices. However, precise targeting allowed this campaign to remain active.

Advertising platform optimization solutions turned out to be unstable with so little data and had to be disabled when they malfunctioned. Sales were strongly impacted by these interruptions. Finally, the major competitors drove up CPCs during critical sales periods.

Alternative customer acquisition channels, relational marketing

Advertising campaigns on Facebook, initiating relational marketing or for retargeting, did not bring significant results despite interesting and relevant content being promoted. Google Ads Display retargeting is effective and somewhat offsets the initial customer acquisition cost. On the other hand, similar audiences were not effective, probably because more data was needed for an ideal prospect profile to be generated.

A sponsorship and loyalty system, in addition to shipments of samples, were implemented, because 20% of customers reorder 12 weeks after the first purchase. These systems were relatively effective but the customer base on which they acted remained limited.

Results obtained with Adsellum’s system

With the implementation of Adsellum’s system on the SEA campaign, the CPA dropped to around €25 while the average order increased to €180 after a few weeks. These results were achieved despite the system tracking CPA rather than ROAS. This means that the system was indirectly promoting advertising content that led to larger orders.

Over the long term, the average order continued to increase and then oscillated around €200, while the CPA remained above €20. Results then fluctuated somewhat around these values. Traffic collapses below €20, so the system was configured to target a slightly higher value. The prolonged improvement in performance was due to campaigns being overhauled further with a few new features and most importantly, Adsellum’s system.

Metric Before After 3 months Longer term
CPA in € 50 25 20 to 25
Average order in € 105 180 About 200
ROAS 2.1 7.2 Up to 10

Cash flow became more stable with more consistent sales, in contrast to the instability experienced with big data solutions that are not adapted to smaller datasets. Additionally, purchases diversified across a wider range of products, rather than focusing on the shop’s specialized and uncommon products.

In this situation, Adsellum’s system targets a broader audience, generates more traffic with a smaller budget, and may lead to some secondary metrics, such as conversion rate, deteriorating. Some intermediate indicators, which are not important in this case, such as CPC, began to fluctuate around better values. However, profitability indicators and other important financial ratios improved significantly, and sales volume continued to increase.

In parallel, proportionally good results were obtained for the Shopping campaign. Also, Display campaigns CPCs were significantly lowered.

Next phase

Reinvesting revenues accelerated customer acquisition, creating a virtuous cycle. However, a new and sudden product price increase temporarily halted this good momentum by lowering dramatically the number of orders. Adsellum’s system adapted and contributed to restoring a stable and profitable sales volume within a few months. This stable foundation freed time to work on referencing and organic traffic.

The SEA campaign expanded further, ultimately reaching approximately 1,500 to 2,300 active keywords depending on the season. These keywords were divided into approximately 150 ad groups, which were selectively disabled when the corresponding organic traffic made them redundant.

Conclusions

Adsellum’s system made it possible to achieve the following using small data samples:

  • Rapidly achieve customer acquisition goals despite poor initial SEO
  • Make sales more consistent, simplifying cash flow management and logistics
  • Increase the average order value and reduce the cost of customer acquisition despite tracking and optimizing CPA rather than ROAS
  • Confidently reinvest the freed margin to accelerate faster