Sri Lanka's Advertising Services: Activity Boundaries, Historical Scale and Delivery Requirements

Sri Lanka’s advertising services extend across strategy, creative work, production, placement and measurement. Official statistics combine part of that work with market research, while agencies sell bundles that cross statistical boundaries. The result is an identifiable service chain with an incomplete measure of its current scale.

The strongest historical benchmark located is DCS Division 73 for reference year 2018: 209 in-scope establishments, 8,165 persons engaged and Rs28.197 billion of output. The figure describes advertising and market research, under a survey frame excluding service establishments with fewer than five persons. It is not a current agency census or the amount advertisers spent buying media.

The central finding: a broader offer than an advertising-spend total

The UN’s ISIC Rev.4 advertising class includes advice, creative services, advertising-material production, campaign placement, media planning and buying, and media representation. It separately excludes stand-alone market research, public relations, commercial-message production, graphic design and photography to other classes. A firm’s offer can combine these activities; its name or network affiliation does not identify a single statistical activity.

This audit examines paid creation and realisation of advertising campaigns targeting Sri Lankan audiences. Media-buying-only work, direct inventory sales, stand-alone PR, research and production-only services remain adjacent activities, even when an integrated supplier offers them. The media-buying audit examines the buying service and associated inventory transaction. Neither report supplies an amount that can be added to the other to create a larger market total.

Evidence searched through 20 October 2026. Coverage is limited: historical establishment indicators, observable service offers and dated rules are supported. Current national agency revenue, shares, prices, profitability and representative client or worker outcomes remain unmeasured here.

The historical division expanded, but the comparison is not a 2026 forecast

Sources are Table 5.2 of the 2017 survey, referring to 2016, and the 2019 survey, referring to 2018. Publication followed in August 2019 and August 2022. Reading the survey title as the observation year would misdate both snapshots.

Persons engaged rose faster than nominal output between these observations. That is a compositional result for the reported division, not proof of weaker employee productivity or wages. Persons engaged includes working proprietors and unpaid family workers; salaries and wages include employer contributions and other benefits. The snapshots are not a matched panel, and neither firm entry nor activity mix can be held constant. Both service frames start at five persons, with full enumeration at 50 persons and sampling below that. Financial-year returns may substitute for calendar-year returns.

The official value-added identity reconciles exactly in each year. The faster increase in intermediate consumption does not reveal agency profit margins: output and intermediate inputs are statistical production measures. The data do not provide the service-only disaggregation, current prices or financial statements needed for a national profitability assessment.

What suppliers actually say they offer

The following comparison uses two first-party offer pages inspected on 20 October. Publicis was retained because it featured in the prior article; Ogilvy provides a second observable provider offering both services and media buying. This selection illustrates activity overlap. It is not a representative agency sample, ranking or company assessment.

The sources are Publicis services and Ogilvy services; neither page supplies a publication date. Advertised capabilities do not establish staffing, sales, contract performance or client satisfaction. They do, however, show why comparing companies by gross billings can mislead: different bundles can contain different quantities of purchased media and subcontracted production.

The brief determines the relevant alternative

A buyer needing a complete campaign is purchasing a different scope from a buyer needing creative adaptation, a single production job or placement of finished material. These task-based segments are analytical definitions; no representative procurement study establishes their national shares or purchasing preferences.

The source-backed offer map supports a practical comparison of scope. An integrated provider advertises several functions together. Specialist suppliers are distinct activities in ISIC, and a platform billing account is another documented route for buying inventory. The evidence does not establish that each alternative substitutes perfectly for the others or identify their local price differences. Work brought in-house also needs staff and tools; no cost panel permits a savings claim.

An observed buyer brief makes the distinction concrete. The Development Lotteries Board’s 13 March 2026 invitation, DLB/PRO/2026/05, sought one year of creative advertising services. Its scope included concepts and production, artwork, periodic brand review, six-monthly focus-group research and digital reporting. It required designated account staff, client service certificates and a creative presentation; the conditions retained DLB approval of content and ownership of production materials. This is a published procurement specification, not evidence that a contract was awarded or that any supplier met the requirements. It shows one public buyer asking for continuing service and evidence of delivery rather than media placement alone.

This matters for competition and customer value. A low media fee cannot be compared directly with a price including strategy, original creative production and continuing measurement. The observable comparison should identify deliverables, rights, revision scope, approval responsibilities, bought-in inputs and the outcome test. The DLB brief supports those concerns for one procurement; it cannot establish private-sector preferences or a national service standard.

A disclosed campaign reveals coordination, not measured superiority

Ogilvy Media’s Sprite case describes weather-responsive LED advertising at eight Colombo locations. Its account identifies backend work by ADA Digital Singapore, integration by Emerging Media and creative by Phoenix Ogilvy. The case page is undated and self-reported; its performance language is not treated as verified outcome evidence.

The useful observation is the division of work. Creative, a data-responsive delivery system and physical inventory had to fit together. This qualifies a simple substitution story in which an advertiser can replace every service with a buying interface. Some disclosed execution depends on several specialised inputs. Conversely, one campaign does not establish widespread adoption, barriers to entry, returns on automation or a permanent advantage for an agency network.

The next evidence test is a dated brief-to-delivery record, with agreed output, supplier responsibilities, cost and an outcome measure whose denominator is known. Without it, public technology menus establish availability claims rather than adoption or effectiveness.

Campaign obligations differ by product, transaction and date

The Consumer Affairs Authority Act, sections 30–31, addresses misleading or deceptive business conduct and specified false representations about goods, services and affiliations. Direction 91, dated 17 May 2023, adds requirements for defined e-commerce entities and platform operators marketing and selling goods, including offer transparency and restrictions on false or hired reviews. Its scope should not be extended to every advertising activity without checking the definition.

Category rules can change what service is legally deliverable. Section 35 of the NATA Act prohibits publishing tobacco or alcohol advertisements, with specified exceptions; section 36 restricts associated sponsorship. The 2015 amendment inspected changes packaging provisions, not those advertising sections. The NMRA’s published medicines guidance requires prior written approval for advertising medicines, devices and borderline products. These sources establish constraints, not the industry’s compliance rate or the outcome of any individual campaign.

Data protection has a dated transition. Gazette 2498/16, issued on 22 July 2026, appoints 1 January 2027 for sections 2 and 3, Part I and Part III of the PDPA. The 2025 amendment permits staged commencement. The order does not, by itself, activate the whole Act, all promotional-message provisions or every penalty. Agencies and advertisers need to identify the provisions relevant to their actual audience-data and supplier arrangements rather than assume a single universal rule.

The IRD VAT guidance retains the standard rate of 18% from January 2024. Its SSCL guidance states 2.5% on liable turnover, subject to the applicable definitions and thresholds. These are not additive estimates of every agency’s effective tax burden. The VAT page also lists a June 2026 amendment; this edition does not provide a transaction-specific analysis of cross-border electronic services.

Billing scale does not establish economic strength

The service chain separates the advertiser’s total budget, inventory payments, agency remuneration and subcontracted inputs. Gross billings can include money that passes to a media owner or another supplier. Net media expenditure and retained service revenue therefore answer different questions. The ONS/Destatis measurement paper provides the conceptual distinction; no Sri Lankan contract panel was inspected to establish one universal accounting treatment.

A supplier with a broad service menu may coordinate more functions, but that alone reveals neither margin nor cash risk. Those require comparable accounts, client concentration and payment terms. No national concentration estimate or numerical competitive score is assigned. Missing commercial evidence remains a limit on the conclusions, not proof that competition or profitability is weak.

The local brand remains an illustration

Publicis Groupe Sri Lanka’s Leo Burnett page presents the local brand and work. That supports its inclusion as an illustrative participant. It does not support prescriptions about how many people to hire, internal capability gaps, account economics or investment requirements. The prior article’s withdrawal of such private-operating-model claims remains in place.

Outlook and market requirements

The assessment date is 20 October 2026, with an outlook to 20 April 2027 and earlier reassessment if rules or material evidence change. The outlook concerns delivery conditions. If briefs combine creative, data, production and media, coordination and a clear allocation of responsibility remain necessary. If a buyer contracts for only one component, the provider’s wider menu may have little value for that job. January 2027’s scheduled provisions create a dated monitoring point for data-related work; further orders or amendments could change it. No spend forecast is inferred from these conditions.

Three proposed market success conditions follow: define the deliverable and its evidence of completion; coordinate the necessary specialist inputs without counting the same cost twice; and substantiate claims and obtain applicable category approvals. Their observable tests are accepted work records, a reconciled supplier/cost map, and dated evidence and approvals for affected claims. These conditions connect the documented service offers, transaction chain and rules. They do not establish any named company’s competence. A requirement changes when the brief, supplier arrangement or governing rule changes.

Methods, limits and publication history

The DCS comparison uses unrounded rupees from Table 5.2, PDF p.31 in the 2017 report and p.19 in the 2019 report. Percentage differences equal (2018 / 2016 − 1) × 100, rounded to one decimal. Monetary levels are nominal. No deflator, exchange-rate conversion or interpolation is used. Both reports describe establishment estimates with sampled smaller units and output-based disaggregation of enterprise returns; there is no matched firm panel here.

The reproducible search covered DCS annual surveys and service-survey documents, CBSL advertising-expenditure queries, classifications, regulator repositories and named provider pages. No current national series reconciling net media expenditure and advertising-service revenue was located through 20 October. The 2018 survey landing page did not expose a usable report, and targeted retrieval of the SLSIC detail failed. The report therefore uses inspected UNSD definitions without asserting an uninspected Sri Lankan subcode. No interviews, confidential accounts or representative campaign, buyer or worker dataset was obtained. Search incompleteness is not proof that no private measurement exists.

The original publication date remains 23 June 2025. The August 2026 correction removed the unsupported US$354 million total, 37% digital share, agency rankings and forecast. Its arithmetic correction remains: US$254 million is 71.8% of US$354 million; if US$254 million were 37%, the implied total would be about US$686.5 million. A US$254 million-to-US$308 million path over four years implies about 4.9% annual growth, not 6.64%. These calculations expose inconsistency; they do not validate any of the amounts. The corrected standard VAT rate and removal of private-agency prescriptions are preserved.

This edition adds the DCS reference-2018 benchmark and service map, replaces absolute statements about absent national data with the bounded search result, and narrows the Online Safety discussion to the dated parliamentary event: the inspected 23 June 2026 record reports second reading and referral to committee. It is not treated as a complete October legal-status check. Further research should seek comparable service revenue and campaign outcomes before making market-share, margin or performance claims.