Case Study Summary
Client: Malaysian professional services group, anonymised. Monthly ad spend: RM 75,000. Engagement start: Q1 2025. Result: CPA reduced 40% in five months; qualified lead volume increased 28% with no increase to budget.
The core problem: Conversion tracking was recording every website form submission including enquiries, subscription requests and internal test submissions as a conversion. The account’s Smart Bidding strategy was optimising toward junk data. The actual cost per qualified lead was more than double what the account reported.
- Step 1: Rebuild conversion tracking to record only CRM-verified qualified leads
- Step 2: Restructure campaigns by commercial intent tier, separating decision-stage from research-stage keywords
- Step 3: Build negative keyword lists from 90 days of search terms data
- Step 4: Switch to Target CPA bidding once 30 genuine conversions per month confirmed
- Step 5: Improve landing page load time from 4.8 seconds to 1.9 seconds on mobile
Introduction
Google’s own guidance on this point is direct. Google Ads conversion measurement documentation states that conversions provide insight into campaign performance and that Google Ads uses conversion data to help you understand ROI and make better-informed decisions about ad spend. When the conversion data is wrong, every optimisation built on top of it is wrong too.
The Starting Condition: What the Account Looked Like at Engagement Start
The client was a Kuala Lumpur-based professional services group with three service lines: corporate legal advisory, HR consulting and compliance training. They had been running Google Ads with a monthly budget of RM 75,000 for approximately 14 months before engaging MYSense. The account was managed by an internal digital marketing executive who also covered social media and email. Google Ads was one of several responsibilities, not the primary one.
The account reported a cost per conversion of RM 480 and approximately 156 conversions per month. To a finance team reviewing these numbers, the account appeared to be performing: RM 75,000 budget, 156 conversions, RM 480 cost per conversion. The problem was that the word ‘conversion’ had never been defined correctly. The conversion tracking tag was firing on every form submission across the website, including newsletter sign-up forms, a ‘request a brochure’ form for the compliance training course catalogue, a webinar registration form and a general contact form. Qualified sales leads, meaning enquiries from businesses seeking legal, HR or compliance services, represented a small fraction of the total.
When MYSense matched the CRM data against the Google Ads conversion records for the three months prior to engagement, the actual number of CRM-entered qualified leads attributable to Google Ads was averaging 66 per month, not 156. The actual cost per qualified lead was RM 1,140, not RM 480. The account was spending RM 75,000 per month to produce 66 qualified leads at RM 1,140 each, while believing it was producing 156 at RM 480 each. The Smart Bidding strategy the account was running, Target CPA set to RM 480, was optimising the bidding algorithm toward a target that did not represent a genuine commercial conversion.
Step 1: Rebuild Conversion Tracking Before Touching Anything Else
The first month of the engagement involved no campaign changes. The entire focus was on making the measurement correct. MYSense worked with the client’s IT team to implement a new conversion tracking structure with three separate conversion actions: a qualified enquiry form submission, filtered to exclude the webinar, brochure and newsletter forms; a phone call of more than 90 seconds from a Google Ads click; and a CRM entry confirmation, implemented via an offline conversion import from the client’s Salesforce instance.
The webinar and brochure form submissions were recategorised as secondary conversion actions, set to observation only, so they would appear in reporting without being used for Smart Bidding optimisation. This distinction, primary versus secondary conversion actions, is one of the most commonly misconfigured elements of a Google Ads account. When every form submission is a primary conversion, Smart Bidding treats them all as equally valuable and optimises indiscriminately across all of them.
The new tracking went live in week three of the engagement. The account immediately began recording fewer conversions than before: 61 in the first full month under the new tracking, compared to the reported 156. The client’s internal team, seeing conversion volume fall by 60%, needed reassurance that this was not a campaign performance collapse but rather an accurate measurement for the first time. Managing this expectation is as important as the technical work itself.
Step 2: Restructure Campaigns by Commercial Intent
With accurate conversion data beginning to accumulate, MYSense rebuilt the campaign architecture. The existing structure had three campaigns, one per service line, each running broad match keywords across a mix of commercial and informational intent terms. A legal advisory campaign was serving ads against searches for ‘what is corporate governance’ alongside searches for ‘corporate legal advisor KL’. These two queries represent fundamentally different user intents and should not compete for the same budget at the same bid level.
The new structure separated each service line into two campaign tiers. The decision tier targeted high-intent commercial keywords such as ‘corporate legal services Malaysia’, ‘HR consultant Kuala Lumpur’ and ‘compliance training corporate Malaysia’, using phrase match and exact match keywords with Target CPA bidding. The consideration tier targeted research-stage keywords using a lower daily budget and Maximise Clicks bidding, with remarketing audiences excluded to avoid paying twice for users already in the sales pipeline.
The campaign restructure also introduced geographic bid adjustments based on branch location data. The client had offices in KL Sentral and Cyberjaya. Users within five kilometres of each office converted at a significantly higher rate than users in the Klang Valley generally. Bid adjustments of plus 20% were applied to searches originating near office locations, directing more budget toward the highest-converting geographic segments.
For corporate teams that want to understand whether their current campaign structure separates commercial intent correctly, MYSense provides a campaign architecture review as part of its Google Ads management services for enterprise accounts in Malaysia.
Step 3: Negative Keywords and Search Terms Governance
The search terms report covering the 90 days before engagement revealed the full cost of running without a negative keyword list. The legal advisory campaign was triggering ads against searches including ‘legal advice free Malaysia’, ‘pro bono lawyer KL’, ‘legal aid society Malaysia’ and ‘consumer protection act explanation’. None of these searches represent a prospect for a corporate legal advisory firm. Each click cost between RM 8 and RM 24.
MYSense built a negative keyword list of 340 terms from the search terms data and applied it across all campaigns. The list covered free, pro bono and legal aid variants across all service lines; student-oriented searches; government and public sector searches that would not convert to commercial advisory clients; and competitor brand names that the campaigns were inadvertently bidding against.
A search terms review cadence was established: the account manager would review new search terms on a fortnightly basis and add negatives for any terms that had accrued spend without converting. This is maintenance work, not a one-time fix. An account that is not actively managed for search term quality will rebuild a wasted spend problem within three to six months regardless of how well the initial cleanup was done.
Step 4: Landing Page Speed as a Campaign Priority
The audit conducted at engagement start included a PageSpeed Insights review of all destination URLs used in the account. The primary landing page for the legal advisory campaign was loading in 4.8 seconds on mobile. Two other service pages were above 4 seconds. Google’s Quality Score system was rating the landing page experience component of these campaigns as below average, which was suppressing Ad Rank and increasing the cost per click relative to competitors with faster pages.
MYSense identified three causes: a 2.3MB hero image on the legal advisory page that had not been compressed; a third-party live chat widget loading synchronously and blocking page rendering; and a Google Fonts import loading six typeface variants when only two were used on the page. Resolving these three issues required one development sprint. Mobile load time fell to 1.9 seconds on the primary landing page and below 2.5 seconds on all other destination URLs.
The effect on Quality Score was visible within three weeks. The legal advisory campaign’s landing page experience status moved from below average to above average. Average CPC on the campaign’s priority keywords fell by 14% in the following month without any bid adjustment, reflecting the improved Ad Rank from the higher Quality Score.
Step 5: Switch to Target CPA Bidding
By month three of the engagement, the account was recording 71 qualified conversions per month under the new tracking definition. This crossed Google’s recommended minimum of 30 conversions per month required for Smart Bidding to calibrate reliably. MYSense switched the decision-tier campaigns from manual CPC bidding to Target CPA, with the target set at RM 750 per qualified lead, reflecting a 30% reduction from the actual CPA measured at engagement start.
The Target CPA target was set conservatively rather than aggressively. A target set too far below the actual account CPA triggers the bidding algorithm to restrict impressions heavily in an attempt to find only the cheapest conversions, often resulting in volume collapse. The correct approach is to set the initial target at or slightly below the current actual CPA and reduce it gradually as the algorithm accumulates data and the account quality improves.
In month four, the Target CPA was reduced to RM 720. In month five, it reached RM 684, which represented the 40% reduction from the engagement start actual CPA of RM 1,140. Qualified lead volume in month five was 85, a 28% increase from the 66 per month baseline at engagement start.
Table 1: Before and after performance metrics for the MYSense Google Ads management engagement, Malaysian professional services group.
|
Metric |
Value |
Note |
|
Reported CPA (before) |
RM 480 per conversion |
Based on all form submissions including non-qualified |
|
Actual CPA (before) |
RM 1,140 per qualified lead |
Based on CRM-verified qualified leads only |
|
Reported CPA (month 5) |
RM 288 per conversion |
Tracking now records CRM-verified leads only |
|
Actual CPA (month 5) |
RM 684 per qualified lead |
Same definition, reflecting genuine 40% reduction |
|
Qualified lead volume (before) |
Avg. 66 per month |
Based on CRM data over prior 3 months |
|
Qualified lead volume (month 5) |
Avg. 85 per month |
28% increase; same RM 75,000 budget |
|
Mobile page load (before) |
4.8 seconds LCP |
Measured via PageSpeed Insights |
|
Mobile page load (month 5) |
1.9 seconds LCP |
After image compression and script removal |
Organisations that want to understand whether their current conversion tracking is recording genuine commercial conversions or inflated non-qualified activity can request an account health check from MYSense as a Google Ads agency in Malaysia serving corporate accounts
Step 5: Switch to Target CPA Bidding
The 40% CPA reduction in this engagement came from five specific, sequential actions. None of them involved increasing the budget. None of them required new creative assets. None of them were the result of finding a better keyword that no competitor had discovered. They were the result of fixing what was broken in the right order.
The sequence matters more than any individual action
Fixing conversion tracking before restructuring campaigns meant the new campaign structure had accurate data to optimise against from day one. Restructuring campaigns before building negative keyword lists meant the negative keyword work was done on a structure worth protecting. Fixing landing page speed before switching to Target CPA meant the Smart Bidding algorithm had both accurate conversion signals and competitive Quality Scores to work with. Had these steps been done in a different order, or simultaneously, the results would have been slower and less clear.
The measurement problem is more common than the management problem
Most Google Ads accounts in Malaysia that are underperforming are not underperforming because of poor creative or inadequate keyword strategy. They are underperforming because the measurement is wrong and the bidding algorithm is optimising toward the wrong signal. Fixing the measurement does not improve the account by itself, but it is the prerequisite for every other improvement. An account running Smart Bidding on inaccurate conversion data is not underperforming because of Smart Bidding. It is underperforming because Smart Bidding is given the wrong instructions.
Frequently Asked Questions About Google Ads Management in Malaysia
In this engagement, meaningful CPA improvement was visible by month three and the headline 40% reduction was achieved by month five. Accounts with more severe structural problems or lower monthly conversion volumes may take longer because Smart Bidding requires a minimum number of genuine conversions to calibrate correctly. An account generating fewer than 30 conversions per month will take longer to see reliable Smart Bidding performance than one generating 60 or more.
Smart Bidding strategies use your conversion data to decide which auctions to bid aggressively on and which to pass. If the data includes non-commercial form submissions alongside genuine qualified leads, the algorithm cannot distinguish between them and will spend budget pursuing both equally. Rebuilding tracking to record only commercially valuable actions gives the algorithm accurate instructions. Without accurate instructions, every downstream optimisation is built on a false foundation.
In Google Ads, a primary conversion action is one that Smart Bidding uses to optimise bids. A secondary conversion action is recorded for reporting purposes but does not influence bidding. Setting newsletter sign-ups, brochure requests and webinar registrations as secondary conversions means the account tracks them without directing the bidding algorithm to pursue them. This is a standard configuration that many accounts never implement correctly, resulting in Smart Bidding optimising toward low-value micro-conversions.
Negative keywords prevent ads from showing against searches that will not convert. Every click on a non-converting search term is budget spent without return. In this engagement, 340 negative keywords removed a significant portion of spend that had been wasted on free legal advice, student searches and public sector queries. The budget saved was redirected automatically toward higher-intent terms by the Smart Bidding algorithm, improving both volume and CPA simultaneously.
Yes, through two mechanisms. First, page speed is a component of Google’s Quality Score landing page experience rating, which influences Ad Rank and therefore the CPC you pay. A faster page reduces CPC by improving Quality Score, independent of what the page contains. Second, a faster page converts at a higher rate than a slow one because fewer users exit before the page finishes loading. In this engagement, improving mobile load time from 4.8 to 1.9 seconds contributed to a 14% reduction in average CPC within three weeks.
The Work That Produced 40% CPA Reduction Was Unglamorous and Sequential
The actions that cut cost per qualified lead by 40% in this account were not creative breakthroughs or algorithmic innovations. They were fixing a broken conversion tag, separating campaigns by intent, building a negative keyword list, compressing images and removing a chat widget. These are the fundamentals of competent Google Ads management in Malaysia. They are also the items most frequently skipped in accounts managed by teams with divided attention or by agencies that do not audit what they inherit.
Corporate buyers evaluating Google Ads agencies in Malaysia should ask specifically what was wrong with their last client account at audit, what sequence they used to fix it, and what the measured outcome was at each stage. The answer to those questions tells you more about agency capability than any presentation about strategy or methodology.
MYSense works with corporate and enterprise accounts across Malaysia to audit, fix and manage Google Ads programmes tied to commercially verified conversion outcomes. To discuss what an audit of your current account would find, contact the MYSense team.
