A recent PPC review, for a furniture e-commerce site, identified some common issues that will help anyone optimising PPC for sales.
Recently, I have been approached by a really interesting industrial furniture retailer. Their (common) concern was that – although traffic was getting to the site – sales were not forthcoming. What more could be done to optimise PPC for Sales?
After a quick chat to get an overview of their business, I requested access to their AdWords account and started to see in detail how the PPC campaign was performing.
Overall the account was well structured and had an extensive list of relevant keywords. The Ads looked good and, where possible, linked to the right landing pages. Conversion reporting was set up and active.
A few things quickly became apparent:
- The objective of the campaign was to drive people to the website or category pages, not to specific products.
- The shopping campaign was currently suspended.
- Search terms were on the whole broad match – certainly work could be done on transitioning these to exact (better relevance, higher ad position and lower cost).
- The negative search terms needed to be expanded, for example, phrases like “gum tree” and “John Lewis” were costing budget. Also, although the site services the UK, location terms tend to indicate that people are looking for local suppliers, so “London”, “Glasgow”, “Scotland” were unlikely to result in sales (as the client is based in the North West).
- The bidding strategy was manual with enhancements (that allows the cost to increase for keywords more likely to convert). Overall the average CPC was 48p and the average position 2.4. This has only delivered 10 conversions at an average cost of £235 CPA (0.2%).
Optimising PPC for Sales: Recommendations
- Some of the ad groups have performed ok. One had a cost of £214 and had generated 4 sales (£53 each) which could be improved – but was much better than other groups. So, the first recommendation was to pause poorly performing groups and increase the bids on better performing / higher priced items.
- Turn off the remarketing. Although not a big cost, the remarketing campaign had generated no sales at all since it started. Targeting had not been particularly refined. Remarketing is great for increasing brand awareness over time. However, display advertising is notoriously poor at generating short term sales. (Although I did notice some interesting placements that could be tested in the future).
- Add new ad groups and keywords. Be more specific and expand the keywords, use more keywords that appear on the website. Some items on the site weren’t being promoted at all because of keyword restrictions.
- Re-test the shopping campaign and optimise the bidding. The shopping campaign had resulted in a couple of conversions, and the cost per acquisition could be improved.
Next Steps in Optimising PPC for Sales
On completion of the analysis and initial report, we agreed that for the next 2 weeks I work on optimising the PPC for e-commerce.
2 weeks isn’t much time to generate sales in such a niche sector. However, the changes made should lead to a better chance of conversion and a return on investment.