
An ad agency is a firm hired to plan and run paid advertising for a client; a digital agency does that work across search, social and the web. The decision that matters for most brand owners is narrower: is the shop accountable for what happens after the click, or only for the campaign it delivered? Performance Marketing World (PMW) took a public position on that question in a page published on 9 November 2026, describing its editorial purpose as championing outcome-based marketing and helping brands and ad agencies get more from their marketing investments.
Ask the plain question first, because the ten pages Google returned for "ad agencies" on 9 November 2026 mostly listed shops and none defined the job. An ad agency sits between a brand and the platforms that sell attention: Google, Meta, TikTok, publishers and, increasingly, retailer media networks. The work splits into four functions.
The first is strategy and planning: choosing which audiences matter, which channels carry the message and what a realistic budget looks like. The second is creative and production. The third is media buying, where the agency places spend and manages bids. The fourth is measurement and reporting, which is where the category now divides.
PMW's own framing, on the page it uses to sell its subscriptions, names the divide directly: it says it champions outcome-based marketing, helping brands and agencies identify best practice, solve problems and get the most out of their marketing investments. The phrase is doing real work. A campaign-only agency can satisfy the first three functions and never touch the fourth in a way the client can audit.

The two labels overlap more than the job titles suggest. The work behind ad agencies historically covered everything from television to print to billboards, while digital agencies grew up inside search, social and web development. In Singapore, most shops that sell to small and mid-sized businesses have merged the two and will answer to either name.
What separates them in practice is not the medium but the workflow. A digital agencies usually owns the website, the landing pages and the tracking setup, which is why it can see what happens after the click and often reports on it. A traditional ad agencies may buy the media and hand the measurement back to the client's own analytics team. Neither arrangement is wrong; not knowing which one you have bought is.
This is the question worth asking in the first meeting. Ask the agency to show a recent client report with the client name removed, and look at the last page. If it stops at impressions, clicks, click-through rate and cost per thousand, the agency is being paid for delivery. If it reaches cost per qualified lead and then revenue, it is being paid for an outcome, or at least measures one.
Outcome-based marketing as PMW frames it is not a new channel or a new tool. It is a claim about accountability: that marketing should be evaluated by the commercial result it produces rather than by the volume of activity it generates. The claim is easy to make in a pitch. It is much harder to hold once a client asks for the definition of the outcome in writing.
Consider a business that sells a service requiring a site visit or a consultation. A click tells you someone pressed a link. It does not tell you whether they had a budget, whether they were in the service area, or whether the sales desk reached them before a competitor did. The four things that decide the sale happen outside the ad platform: the form, the qualification call, the booking and the follow-up after a quotation.
An agency that cannot see those four steps is guessing at which campaigns work. An agency that can see them can reallocate budget from the campaigns that produce enquiries nothing converts to. The technical name for the gap is attribution, and the practical test is whether the report can name the deal.
Three things change once the outcome is defined, and each one has a cost attached.
The first is the denominator. Cost per lead means nothing until everyone agrees what counts as a lead. A form fill, a booked call and a sales-qualified conversation are three different events with very different values, and shrinking the definition quietly improves the reported number.
The second is the feedback loop. If qualifying conversions can be passed back to the ad platform, smart bidding can optimise for them instead of for clicks. If they cannot, the platform optimises for what it can see, which is usually the click.
The third is the exit clause. An agency willing to state a target in public, measured one way, and to let the client leave if the target is missed, is taking a different kind of risk from a shop paid on a monthly retainer regardless of result.
| Question to ask | Campaign-only answer | Outcome-based answer |
|---|---|---|
| What is the reported end point? | Impressions, clicks, CTR | Qualified lead, closed-deal value |
| Who defines a qualified lead? | Not stated | Written down and agreed before launch |
| Can conversions go back to ad platforms? | Not usually | Yes, via offline or server-side upload |
| How is the fee structured? | Retainer or percentage of spend | Stated per plan, with the measure named |
| What happens if the target is missed? | Contract runs | Client can exit under a stated condition |
| Whose data and account? | Agency account | Client's own account and tenant |
The table is not a verdict on any single agency. A campaign-only shop can be the right choice for a brand that already has a measurement team and simply needs reach. It is the wrong choice for a brand that has no way to tell whether the spend worked.

The PMW page on 9 November 2026 is itself a useful case in buying a decision input. It sells two memberships. The Information tier costs £24 per month. The Knowledge tier costs £49 per month and adds Campaignlive.co.uk access, deeper survey data, an in-housing report and a quarterly AI global agency league table.
A team licence starts at 3+ users and is quoted as bespoke. On Haymarket's subscription shop, the same publication's business subscriptions are priced by region and term: a three-month subscription by direct debit is £82.00 and a one-year subscription by direct debit is £296.00 in the United Kingdom, while a three-month subscription outside the UK and Europe is $118.00 and a one-year rest-of-world subscription is $429.00.
Two practical details sit on the subscription page. Online access begins within 24 hours of payment being processed, and the top tier includes The Performance Power 100, a ranking of the top 100 most influential people in the performance marketing industry worldwide. The shop separately notes that team discounts apply for 5+ users.
For a brand owner in Singapore, the arithmetic is worth doing in SGD before buying. The point of the purchase is not the reading matter; it is the standard the industry publishes, which you can then hold an agency to. A £49 monthly membership is a modest sum next to a media budget, and it is the cheapest way to learn what a credible report should contain.
Industry knowledge and account-level evidence are different goods, and the page is honest about being the former. The Performance Power 100 ranks influence; it does not rank agencies by client results. The league table tracks new business wins; it does not say whether those accounts produced returns. A brand that buys the subscription and stops there has bought context, not proof.
The proof has to come from the account, which is what the readiness questions below are for.
Before any conversation about fees or terms, check whether the account has the plumbing to support the claim. An agency that promises outcomes on an account with no conversion tracking, no CRM connection and no agreed definition of a qualified lead is promising something it cannot measure.
A short readiness list:
If four of the five are missing, the honest first project is not a campaign. It is the measurement setup. Most agencies will sell the campaign anyway, which is why the question needs to be asked by the client.
Topkee has built its Singapore offering for brands working with digital agencies around the same gap PMW writes about. Its position is that brands keep spending on Google, Meta and TikTok without being able to prove return, and that the cause is rarely the ad platform but the separation between systems: site, CRM, chat, e-commerce and ads hold different data and no loop closes. Topkee Cloud connects existing CRM, ERP, e-commerce or legacy databases, and qualifying conversions are passed back to Google and Meta through offline conversion upload so that smart bidding optimises for revenue rather than clicks.
ROAS Engine, the paid-media service, measures a single figure: ROAS calculated through to closed deals, with closed-deal value divided by ad spend. It runs on three openly priced plans with monthly billing, no long-term lock-in and no separate platform setup fee; media budget is paid through the client's own ad account, so delivery data stays with the client. The published commitment sets a starting floor and says if the client misses the published ROAS threshold they can leave without penalty, with no-fault cancellation available on 24 hours' notice before the next billing date.
Two supporting systems address the post-click steps. Lead Engine covers capture, qualification, booking and closing, each reported separately, and attaches a numbered, prioritised action list to each period's report. Spend Shield puts the waste audit inside every report rather than selling it as an add-on, listing wasted spend on blank search terms and fatigued creative, with the status of each fix visible in the system. Topkee Growth System is positioned against campaign-only agencies with three stated differences: more post-click visibility, less tool fragmentation and faster optimisation.
These are statements about what the services do, not a claim that outcomes follow automatically. A brand comparing shops should treat them as the spec sheet to check against, and ask any agency, Topkee included, for the same evidence: the measure, the period and the action list.

The advice above assumes a business with a sales process long enough to have a post-click stage. A brand that sells a low-value item straight from a product page can often judge an agency on return on ad spend alone, because the purchase happens on the click.
It also fits less well for a marketing team that already runs its own attribution and simply needs media buying capacity. In that case, a campaign-only shop with a clean hourly rate is a reasonable buy, and the outcome framework adds meetings rather than information.
Finally, a firm with no CRM and no plan to adopt one should not hire an outcome-based agency first. The sequencing is wrong; the measurement project comes before the media.
If you are unsure whether your account can be judged on closed-deal value, start with the audit rather than the campaign. Topkee runs a free read-only account audit with a two-minute form and a diagnosis within three working days, which is a low-cost way to see what the report would look like. For a wider view of how AI-driven reporting and campaign automation change the picture, read how AI-powered tools can boost Google Ads campaign performance and how AI-powered Google Display Ads can lift ROAS.
An ad agency traditionally buys advertising across all media, while a digital agency works inside search, social and web. In practice most shops serving small and mid-sized Singapore businesses do both, so the labels describe history more than current work. Ask what the agency actually runs for its clients.
Ask for a recent report with the client name removed and look at its final page. If the report ends at cost per qualified lead or closed-deal value, the agency measures outcomes. If it ends at clicks and impressions, it measures delivery.
It gives you the industry's own definition of good reporting, which is useful when you are comparing agencies. It does not tell you anything about your own account. A £24 or £49 monthly membership is cheap context; it is not evidence.
Four things: the definition of a qualified lead, the formula for the outcome measure, the reporting period and the condition under which the client can leave. Without the first, the rest are negotiable in the agency's favour.
Long enough for the sales cycle to complete at least once. For a business that closes deals over weeks or months, the first honest read arrives after that cycle, not at the end of the first month.