Zoe Bates takes a closer look at forecasting paid search ROI and offers her best practice advice for creating, monitoring and maximising on your PPC forecast to generate more online revenue.
Most businesses need their forecasting to follow their financial year. It makes sense, as you can also map out seasonality and, over time, work to iron out the creases that can appear in low seasons. A 12-month forecast will enable you to see what factors affect month-on-month variations, as well as showing which channels perform best at what time of year. This is great for budget allocation.
Make sure you include all your paid search activity – have a top line view, but break it down by channel:
Even go so far as to split by Brand and Non Brand and at a product category level - each category may have different targets based on profit margins, delivery costs, return rates etc. Each channel will have a different CPC, CTR and conversion rate, so it is important to see these in isolation.
What Type of Online Retailer Are You?
Are you Pureplay or Omnichannel? If you are Omnichannel what value does your website hold as a research tool to drive in-store sales? Typically, paid search campaigns are used to drive traffic and web sales. However, they can also be used as a powerful channel for driving traffic, and ultimately sales to in-store. The trickiest aspect of this strategy though is its ability to accurately track performance and measure the ROI.
Do you have Reserve and Collect functionality or a voucher system to track in-store sales. Do you take telephone orders? If so, this needs to be tracked in paid search as it’s important to know the keywords driving all sales. Securing budget is hard when you are targeted by ROI and your website is also there to support local stores and call centre sales.
You can also include seasonality (you will want to increase or decrease budget according to your most competitive times of the year – use Google Trends to see this if you don’t have the insight). Also, think about your own promotional activity. It is important to align your budgets to you major sales events to ensure you can make the most of the promotions. Also proportion your budget by month and channel, as our examples below show.
Also layer into the forecast any site improvements. Are you planning on doing A/B split testing, conversion rate optimisation or launching new products, which will impact overall conversion rate or average order value. All of this needs to be layered in to give as realistic a forecast as possible.
5 Considerations For Creating a 12-Month PPC Forecast
Over time, you will be able to map out trends in impressions over 12 months and show how these improve CTR, CPC and CR. Consider what keywords you are going after and their impressions share. But remember, this is always a guesstimate.
Let’s work through a 12-month example. For the purpose of this blog our fictional company stars with an ad spend of around £4,000 and an average order value of £100*. Let’s see how this pans out over a one-year period and work out what the stats are telling us about sales, budget and forecast in terms of:
- Brand Vs non-brand
- Google Shopping
- Total Spend
- Brand Vs Non Brand
Remember, brand keywords will always have a higher CTR over non brand keywords.
Over the year, forecast how keywords can improve CTR through optimisation. Layer in increased CTR for seasons such as Christmas, and split this between brand and non brand.
Here’s an example of how brand and non-brand might compare:
As we can see, brand has a substantial lead over non-brand across the board. Seasonal fluctuations are evident in the build-up to Christmas and January sales, dropping in February with minor fluctuations across the spring and summer months. It is also important to factor in the fact that the more your push non brand the more it may expose the demand for your branded searches. Always monitor this to ensure you have sufficient budget to maximise on impressions and share visibility.
- Google Shopping
Shopping is more competitive and more seasonal. Expect order values to increase and CR to go up in December ahead of Christmas, while for January sales the order value will go down due to sales discounts, but the CR will stay high.
Make sure you consider the quality of your Google Shopping feed. Always have a good number of products showing, and use your feed with custom labels to manage your products at ROI, margin or seasonal availability to ensure control over spend and product visibility.
In 2015/2016 we saw a real shift in paid search budget going towards shopping (60%) and more budget going into feed optimisation. However, this takes time. With omnichannel brands, you also need to consider the power of your site as a research tool to drive in-store sales.
And remember to factor in any new product launches coming later in the year – the more products you enter in the auction, the bigger the budget needed.
Remarketing enables your ads to show to people who have previously visited your website or used your mobile app. For those that visit your site without converting, remarketing is a great way to reconnect with them by showing relevant ads as they browse.
You may want to take this a step further and employ dynamic remarketing, whereby your ads actually push the specific product or services people visited on your site. This can deliver customised, high-performing ads. And consider looking beyond Google Display Networks for your remarketing activities. Have you considered platforms such as Criteo or AdRoll, for example?
Remarketing activities will start after your initial campaign. Impressions and clicks will be lower, but so too are costs and, because the customer is already familiar with your brand, your CR will be higher.
Remarketing is a great tool for maximising on peak conversion periods, such as Christmas, or for pushing promotions. You may even want to create remarketing lists that show ads to visitors who have performed specific actions on your site – such as offering a discount to someone who abandoned their basket, or even to remarket new products or services to those visitors who have purchased from you previously
Are you planning to start implementing regular discounts? Remarketing works best when there is a strong discount call to action, so you may want to push this channel more when you have strong offers.
Microsoft’s search engine Bing is gaining ground in the online advertising arena, now showing on Yahoo and AOL. As such, we recommend Bing should receive 10% of your AdWords budget as a starting point.
You should manage your Bing ads in much the same way as you do your Google AdWords campaigns – with a built-out keyword list, optimised ad creatives and targeting options. Bing has recently added features such as sitelinks, app extensions, shopping ads, remarketing and annotations, rounding out its paid search offering.
Bing Ads have less competition and cheaper CPCs and we also find higher average order values – giving you more room to experiment without gambling with your ad spend.
With a lower percentage of your budget, Bing will cost you less but also bring in less revenue. However, in some cases we have seen that the Bing demographic tends to have a higher average order value. As Bing grows, and as your ad campaigns mature, the potential could be there to expand this market further.
- Total Spend
Once you have optimised your account for brand and non-brand terms across all channels, including Google Shopping, Remarketing and Bing, you will have a comprehensive forecast that should look something like this:
12 Months On – Let’s Go Again!
You’ve completed your first 12-month forecast. The good news is that forecasting gets easier from this point on. Every year, you will learn more about how your account performs, which channels are favourable at which time of year, and what to expect from remarketing campaigns.
The important thing here is not to get complacent. It is easy to carry out a forecast, and then forget about it. For forecasting to work it needs to be an organic tool that is used daily and is allowed to grow. Keep your forecasting up to date, and map it against real-time data to see how accurate you have been with your forward calculations.
Embrace glitches in the forecast – as these are your key learning points. If your account throws you a swerve ball analyse where this came from, what skewed the figures, and what you could have changed to avoid any drop in performance. Then implement these learnings so that, next time, the drop is less.
Want to implement a 12-month paid search forecast but not sure where to start? Download your free 30-page eBook How to Build an e-Commerce ROI Forecast for Paid Search and learn from our best practice modelling techniques.
*The statistical data in this blog is Example Data only. This is for illustration purposes only in describing our forecasting methodologies and should not be taken as a tried and tested business example.