At Hubstaff, we take growth seriously.\nWithout it, we wouldn’t be in the position we are today, with over 3,000 customers and $110k+ in monthly recurring revenue.\nBecause growth is so important to us, we’ve experimented across a number of different channels to see what works best for our business.\nThe results we’ve seen from our campaigns have been interesting, surprising, and sometimes disappointing.\nAs growth is something that every startup relies on to survive, we’d like to share our experiences with growth marketing and paid acquisition with you in the hope that you’ll be able to replicate and build on the successes we’ve had, and learn from some of the areas that haven’t worked well for us.\nThis post contains examples of the campaigns we’ve used, the channels we’ve explored, and the results we’ve generated.\nIs paid user acquisition worth investing in? You might be surprised... Click To Tweet\nLead Generation Campaigns\nAs Hubstaff is a bootstrapped startup, we have to make sure that the money we spend on paid acquisition campaigns generates a positive return: we don’t have millions of dollars in VC money to throw at various different campaigns, so we have to be thoughtful about where we spend our dollars.\nThis means that we need to see clear and measurable returns from all of the campaigns we run.\nGoogle AdWords\nWith this in mind, we turned straight to Google AdWords Search ads as the first channel for our marketing spend.\nOur goal for this campaign was to capture the search intent of our users to better understand what they were looking for. We targeted users that were searching for specific terms (like ‘time tracking software’ and ‘employee monitoring software’) that are related to our niche.\nFrom there, we created individual Ad Groups based on keyword themes and built custom landing pages for each type of ad to link to, so that we achieved a good quality score (QS) and generated leads at the lowest possible cost.\nWe used Google Keyword Planner and UberSuggest to help generate keywords for this campaign.\nFrom the beginning, we set the keyword match types for each of the ad groups to either ‘phrase match’ or ‘exact match’, so that we were able to tighten our targeting and avoid spending ad money on broad match keywords that might not have generated the same result.\nIt’s not always inappropriate to use broad match keywords, because in some campaigns they can help to generate useful information that can help you narrow down search intent: but for this campaign, we knew the exact terms we wanted, so broad match keywords weren’t useful.\nTo help narrow our campaign even further, we also created a list of negative keywords. These tell Google not to show the ad to anyone searching for a specific phrase, such as ‘free’, and help keep the ad targeted at our exact demographic.\nOur Google AdWords Ads\nTo make our ads as engaging as possible, we used dynamic keyword insertion to customise our ads to the individual user’s search. We also used the callout ad extension to allow us to say more about our product.\n\nTargeting and Budgets\nTo make the campaign as successful as possible, we were specific in our targeting so that our ads were only placed in front of the people we knew were our target market. For this campaign, we wanted to target employers and managers of businesses and remote teams, rather than freelancers or consultants.\nWe also chose to target people from regions or countries where we see the majority of conversions from organic leads to paying customers. These target countries were the US, Canada, Australia, and the UK, among others.\nAs you can see from the screenshot below, our target keywords have a high ‘suggested bid’ amount (just think of the daily cost if we wanted to drive lots of traffic!), but that is typically offset by the revenue we get from a paying customer. What’s more, this amount isn’t the actual CPC amount that’s paid when someone clicks on the ad, but it helps to give a rough estimate of the type of cost involved in running a campaign in this niche.\nThe suggested bid amount helps your ad to appear in search results for your target customers. If you enter a bid amount that’s too low, you either won’t appear in search results at all (as all of your competitors will outbid you), or you’ll only appear in front of low-quality leads (and this will end up costing you far more money in the long-run).\n\nCompetitor Analysis\nIf you’re in a competitive niche and want to check out what marketing campaigns your competitors are running (and this is something we absolutely recommend you do) you have two options:\n\nType in the keywords you’ve picked to target in your campaigns, and see which of your competitors is advertising and what their ads look like\nUse specific tools, like SpyFu, to see exactly what ads your competitors are generating and gain an understanding of other factors that relate to their campaigns\n\nWhile option 1 is good for low-level research and can help you to get ideas for your ad campaign, we rely on option 2 to provide us with the information we need to run campaigns successfully.\n\nThe screenshot above shows you some of the information that SpyFu generates as part of their competitor analysis. The tool lets you understand the approximate number of clicks a competitor’s campaign gets each day, what the cost-per-click is, what ad variations they are running, and what positions their ads rank at.\nAll of these bits of information are invaluable when it comes to creating a successful Search Ads campaign, and can be used to tailor and tweak the ad copy that we create.\nFacebook\nFacebook ads are a great way of reaching a large number of people in a cost-effective and efficient way. With Facebook, we wanted to target a specific audience for lead generation purposes.\nWe wanted to target founders, CXOs, and managers of various remote teams, agencies, and startups. These are the people who ultimately have the power to make the decision to use Hubstaff for their teams, so it only makes sense to target them.\nThe CPC for Facebook ad campaigns tends to be much lower than Google AdWords, however, the quality of the leads can vary as you aren’t targeting ads based on a user’s intent but on their overall interests.\nWhat’s more, if your audience targeting is not very precise, the cost per lead can skyrocket as the CPC climbs and the conversion rate drops.\nOur Facebook Ads\nWhen it comes to advertising on Facebook, visuals matter.\nHere are some examples of ad campaigns we’ve run on Facebook – notice how they use bright and bold colors that grab attention.\n\n\n\nWhile Facebook can be a great way to generate low-cost, high-quality leads, we found that this campaign was expensive compared to other solutions.\nRemarketing\nRemarketing is a great way to help increase conversions and get people who have previously visited Hubstaff, or signed up for Hubstaff, to return to the site and complete a new (and more important) action.\nWe decided to focus our remarketing campaigns on two objectives:\n\nRemarketing to visitors of certain pages on the website who have not yet signed up for a free trial\nRemarketing to free trial users who have not yet converted to a paid account\n\nBelow is a diagram that gives an overview of our remarketing strategy.\n\nThere are a variety of remarketing channels that you can use for your campaigns, including AdRoll, ReTargeter, and others. We chose to use Google’s Display Network and Facebook for this campaign, as we’ve previously had good results with both.\nGoogle Display Network\nWe were able to generate a good number of leads at a reasonable cost by running remarketing campaigns on Google’s Display Network. However, we didn’t really see a good free-to-paid conversion rate: this suggests that the leads we generated here were low quality.\nBelow are some examples of ads we ran on Google.\n\nFacebook\nFor users that had already started a free trial, but hadn’t converted to a paid account, we set up a remarketing campaign on Facebook that showcased our customer case studies so that we could provide social proof to push them down the funnel and convert them into paying accounts.\nUnlike the ads run on Google’s Display Network, the remarketing ads on Facebook worked well and generated a few paid account conversions, which was great considering the budget.\nHere is an example of some campaigns we ran:\n\n\nTo make sure that the conversion attribution was properly set up, and so that we had accurate data, we made sure to have the relevant Facebook pixels set up across our site. We also put URL parameters in place for each ad to help with tracking.\nContent Promotion\nAt Hubstaff, we create a huge amount of content each year with a goal of providing value to our customers and visitors. The content we create (like the post you’re reading right now) contains valuable and actionable information that our readers can use to better run their businesses and remote teams.\nTo help amplify the reach our content gets, and to increase the number of people benefitting from our posts, we turned to running content promotion campaigns with native advertising platforms.\nWe ran a campaign with OutBrain with the hope of driving high-quality traffic to our articles and features in the hope that new visitors would spend time on the site, sign up for our newsletter, or download our eBooks.\nUnfortunately, even though we received a tonne of traffic at a very low cost, the quality of this traffic was very poor, and we didn’t see any evidence of strong engagement, or any soft conversions.\nHere’s an example of one of the campaigns we ran on OutBrain\n\nAs you can see, one post out-performed the group significantly and consumed the majority of the ad budget. The average CPC we received for this campaign was around $0.25, which is low, but this cheap cost is completely offset by the quality of the traffic.\nAs a result of the poor performance of this campaign, we’re actively experimenting with Facebook to distribute our content to our target market, hopefully, this will lead to an increase in our newsletter subscriber base and to more free trials and conversions.\nA\/B Testing and Analytics\nA\/B testing can be a great way to spot design flaws in your website, increase conversions, and improve the quality of experience you provide for your visitors.\nFor us, the goal of our A\/B tests was to discover which site variation drove the greater number of conversions.\nAs you can see from the campaign data below, we ran two ads with the Basecamp and Asana integrations as the main feature and ran two variations of each ad.\nWe didn’t have a large budget, so the overall traffic level was low, but the data we received helped us to understand how a user interacts with our pages and where the conversions happen.\nTo track on-page activity, we used the heatmap feature in VWO. We also set up various funnels in Woopra for each of the channels that we spend our advertising budget on. This helps us to track and analyze each step in the conversion funnel and decide which areas we need to optimize and improve.\nHere you can see an example of a Facebook remarketing campaign we ran to convert free trial users into paid accounts:\n\nWith Woopra, we are able to slice and dice the data across the funnel as needed. We also appended campaign\/ad-specific URL parameters and created landing pages for each campaign so that conversion attribution is as straightforward as possible.\nWhat Have We Learned about Paid Acquisition?\nOverall, we’ve found it difficult to make a paid acquisition channel scalable. We’ve found that we can generate leads at reasonable costs, but that converting them into paid accounts is challenging.\nGoing forward, we are looking at ways to improve the conversion rate from free to paid, without significantly increasing our cost per acquisition. We’re going to experiment more with Facebook for driving blog traffic, and continue to refine our approach to other marketing channels.\nHave you experimented with growth marketing? Share your experiences below in the comments!