What Is Manual Bidding?
Manual bidding, formally known as Manual CPC (cost per click), is a bid strategy in Google Ads where the advertiser specifies a maximum bid for each keyword or ad group. When your ad enters an auction, Google uses the bid you have set as the ceiling for what you are willing to pay per click. If another advertiser wins the auction, Google's second-price auction means you often pay less than your stated maximum, but you will never pay more.
Under Manual CPC, you decide the bid for each keyword individually. A keyword like "emergency plumber Johannesburg" might warrant a R60 bid because it signals high purchase intent, while a broader keyword like "plumbing tips" might only justify R8 because users searching that phrase are far less likely to convert. This granularity is the main advantage of manual bidding: you can reflect your actual knowledge of keyword value in every bid, rather than relying on an algorithm's estimate.
Manual CPC is available with an optional enhancement called Enhanced CPC (eCPC). When eCPC is enabled, Google may raise or lower your manual bid by up to 30% in individual auctions where it predicts a higher or lower conversion likelihood. eCPC is a halfway point between fully manual and fully automated bidding, and it can improve results while still anchoring to your manual bid levels. If you prefer zero algorithmic adjustment, you can disable eCPC entirely.
Compared to automated bid strategies like Target CPA or Maximise Conversions, manual bidding requires more active management. You must review keyword performance data regularly, identify which keywords are over- or under-bidding relative to their conversion rates, and adjust bids accordingly. For an experienced Google Ads manager with the time to do this, manual bidding can produce excellent results. For a business owner running their own campaigns while also managing day-to-day operations, automated bidding often delivers better outcomes because it adjusts in real time without requiring constant attention.
When to Use Manual Bidding
Manual bidding is the right choice in several specific scenarios. First, it is the appropriate strategy for any new campaign or account without conversion history. Google's Smart Bidding algorithms need a minimum of 30 to 50 conversions per month per campaign to function reliably. Without that volume, the algorithm has insufficient data to predict conversion likelihood, which leads to erratic bid decisions. Starting with Manual CPC or Maximise Clicks (which is automated but focuses on traffic rather than conversions) allows you to gather data while spending your budget predictably.
Second, manual bidding suits campaigns with very small budgets where every rand matters. If your daily budget is R50 and Google's Smart Bidding system decides to spend R40 on one click it predicts will convert, you have very little budget remaining for the rest of the day. Manual bidding gives you direct control to avoid any single click consuming a disproportionate share of your spend.
Third, manual bidding is valuable when you are testing a new keyword set and want to observe raw click behaviour before applying algorithmic optimisation. Setting all keywords at the same moderate bid level (say, R15 per click) removes bid variance as a variable, letting you compare ad copy performance, landing page conversion rates, and keyword intent signals on equal footing.
Fourth, experienced Google Ads specialists sometimes prefer manual bidding for branded keyword campaigns, where the intent is extremely clear and consistent and there is no need for machine learning to predict conversion probability. For a business defending its own brand name in search, a high manual bid ensures top position at a predictable cost.
Manual Bidding In Practice
A Cape Town law firm launches its first Google Ads campaign targeting people searching for family law services. They have no prior conversion data in Google Ads, so Smart Bidding is not yet viable. They begin with Manual CPC, using Google Keyword Planner to identify that clicks for "divorce attorney Cape Town" cost between R35 and R70 in their market. They set an initial maximum CPC of R45 per click across their core keywords, which sits in the mid-range and gives them sufficient impression share without overpaying.
Over the first six weeks, they accumulate performance data. They observe that "divorce lawyer Cape Town urgent" converts at 12% (one contact form submission per 8 clicks at R45 each, equating to a R375 cost per lead), while "divorce process South Africa" converts at only 1% (one lead per 100 clicks, or R4,500 per lead). They reduce the bid on the informational keyword to R8 and increase the bid on the high-intent keyword to R65. After three months of this iterative adjustment, they have accumulated 45 conversions and switch to Target CPA at R400 per lead, allowing Smart Bidding to take over with sufficient data to operate effectively.
This pattern, starting with manual CPC to collect data and then transitioning to automated bidding once conversion volume is established, is the recommended approach for most South African SMEs entering Google Ads for the first time. It avoids the underperformance of Smart Bidding on thin data while building towards the scale where automation adds genuine value.
FAQ
When should a South African small business use manual bidding instead of smart bidding?
Use manual bidding when your campaign is new and has no conversion history, when you have fewer than 30 conversions per month (too few for Smart Bidding to work reliably), when you want tight budget control on a limited spend, or when testing a new keyword set. Once conversion volume grows, switching to Target CPA or Maximise Conversions typically improves performance.
Does manual bidding perform worse than automated bidding in Google Ads?
Not necessarily. An experienced manager with time to monitor and adjust bids can match or exceed automated performance on smaller accounts. However, Smart Bidding outperforms manual bidding on larger accounts with high conversion volume because machine learning can react to real-time auction signals like device, location, time of day, and browsing behaviour in ways no human manager can match at scale.