Teacher's guide · Experiment 026

Hungry for Clicks

Teams are rival PPC agencies for Nibbo, a new vegan chocolate brand, with £500 of ad money a round. They pick keywords, then a platform and a way to pay, then a bidding strategy just before a chaos card hits. Every round winner’s name shines with gold stars; the overall winner gets the golden wrapper, a trophy and chocolate rain. A final chart shows what happens when the ads are switched off. This guide explains the model, every design assumption, the results to expect and what to click, say and watch for.

15 to 25 minutes2 to 8 teams or solo3 quick roundsOne screen⏱ What affects timing
The idea

What students learn

1. Intent beats volume

"chocolate" has 50,000 searches and eats the budget with browsers. The brand name and the long-tail "buy vegan dark chocolate online" have few searches but bring buyers. Broad match without negatives pays for "chocolate cake recipe"; a "–vegan" negative switches off your own product.

2. Pay for what the goal needs

For an awareness brief, buying clicks (CPC) or using Search, which only reaches people already looking, reaches few people. CPM or CPV on a visual platform reaches many. The same £500 reaches 5,000 or 70,000 people.

3. Smart bidding needs good data

Conversion-based bidding wins, until tracking breaks. Quality fixes (landing page, ad copy, assets) lower cost per click and lift sales; raising bids instead only makes clicks dearer.

4. PPC is a tap, SEO is a tree

Paid traffic arrives the day you pay and stops the day you stop. Organic traffic builds slowly and keeps coming. Most brands need both.

Built for search marketing and digital marketing classes. It is the quick, whole-campaign game; for depth, pair it with Match Point (match types, the search terms report and negatives over three weeks), Bid & Rank (a whole PPC account and its metrics), Ad Auction Live (how Ad Rank and the auction work) and Rank Rescue (SEO). Nibbo and every number are fictional.

Read first

Five golden rules

  1. Reloading is safe.The game remembers where you are in this browser tab. Closing the tab clears it.
  2. Picks can change until you reveal.After a reveal, "Change picks" takes you back to that round if a team was entered wrongly.
  3. Two keywords, one match type, one negative in round 1.Tapping a third keyword drops the first one. If nobody taps a match type it stays on exact; the negative stays on "none".
  4. Each round is scored on its own.Round winners get 10 points, then 7, 5, 4, 3, 2, 1, 1. Points add up; ties on the podium go to round 3 profit.
  5. The chaos card is random.It is drawn when you lock round 3, after every team has chosen, so nobody can play to it.
How it is built

The model

Every round: £500 budget · average order £20 · £10 profit per order before ads
Round 1 clicks = searches × match reach × click-through rate; junk clicks = clicks × match junk share × (1 − negative cut)
Orders = (clicks − junk) × conversion rate × match quality; if spend goes over £500, everything scales down to fit
Round 1 score = orders × £10 − spend (+ £50 earned media for the misspelling)
Round 2 score = people aged 18–34 reached = £500 ÷ price per 1,000 × 1,000 × platform's 18–34 share (× 1.3 for video views)
Round 3: CPC = strategy CPC × fix; conversion rate = strategy rate × fix; then the chaos card; score = orders × £10 − spend

Round 1: six keywords (pick two)

CardSearches a weekClick-throughCPCConversionNote
A. chocolate50,0002%£0.601%Huge, vague: eats the budget
B. vegan chocolate8,0005%£0.8010%The category
C. buy vegan dark chocolate online1,5008%£1.0016%Long-tail, ready to buy
D. chocolate gift box10,0004%£1.206%Busy and pricey
E. nibbo chocolate30030%£0.2020%Brand: cheap, high intent
F. choclate4,0003%£0.101.5%Misspelling stunt: +£50 buzz
Match typeReachJunk share of clicksConversion quality
Exact×10%×1
Phrase×1.825%×0.9
Broad×3.550%×0.75

Negatives: –recipe removes half the junk, –free removes 30%, and –vegan is the trap: it removes nothing useful and blocks keywords B and C completely.

Round 2: platform and pay model

PlatformShare aged 18–34CPMCPC (per 1,000 reached)CPV (per 1,000 reached)Best reach
Google Search40%–£40–5,000 (any choice becomes CPC)
Instagram & Facebook55%£6£15£945,833 (CPM)
TikTok70%£5£18£770,000 (CPM)
YouTube50%£7£25£840,625 (CPV)

Round 3: bidding strategy and one fix

StrategyCPCConversionSpendsSmart?
Manual CPC£0.807%£500No
Maximise clicks£0.655%£500No
Maximise conversions£0.9510%£500Yes
Target ROAS 300%£0.9011%£400Yes
FixCPCConversionQuality Score component
Faster mobile landing page×0.90×1.25Landing page experience
Ad copy matches the keyword×0.85×1.05Ad relevance
Sitelink & price assets×0.92×1.10Expected click-through rate
Skip the fix, raise bids 15%×1.15×1None: same ad, dearer
Chaos cardWhat happensWho is protected
Bot attack15% of clicks are fakeSmart bidding: only 5% lost
Tracking brokeSmart bidding falls to a 7% conversion rate, minus 10%Manual and Maximise clicks don't notice
A rival bids on your keywordsCPC +20%Matching ad copy: only +10%
An AI Overview appears30% fewer clicks (and less spend)Sitelink & price assets: only 20% fewer
Be transparent

Design assumptions and simplifications

The same list appears in the game under "How this game simplifies reality".

  1. Every number is invented for teaching.Searches, click-through rates, prices and conversion rates are round, plausible figures chosen so that good decisions win clearly. Real figures vary by market, season and account.
  2. Match types are three dials.Real matching is decided by Google's systems in real time and also uses your landing pages, other keywords and data, so real broad match can do better or worse than this.
  3. Search ads are sold per click.A CPM or CPV plan on Google Search becomes CPC, as in real Search campaigns. CPM and CPV prices for social and video are illustrative, and pay models vary by platform and objective.
  4. Bidding strategies are four fixed profiles.Real smart bidding learns over time and needs enough conversion data. Enhanced CPC is not offered: Google retired it for Search and Display campaigns from the week of 31 March 2025.
  5. Quality fixes are multipliers.They echo the three Quality Score components (expected click-through rate, ad relevance, landing page experience). Quality Score itself is a diagnostic, not an auction input.
  6. Chaos cards are stylised.Invalid clicks, broken tracking, new competitors and AI Overviews are all real; the percentages are for the game. Google filters invalid clicks out of billing and credits some back; ads can appear above, below or inside AI Overviews.
  7. The "switch the ads off" chart is illustrative,not data. The shape is the point.
  8. Nibbo is fictional.Any resemblance to a real company is accidental.
Worked answers

Results you can predict

Every result is fixed by the picks; only the chaos card is random. Teams don't affect each other's money, only each other's rank.

Round 1 (profit after ads)

PicksProfitWhy
C + E, phrase, –recipe£279Best in the game: long-tail plus brand, with the junk cut
B + E, exact£242Category plus brand, no waste
C + E, exact£234Safe, but small
E + F, exact£218Brand carries it; the misspelling adds buzz, few sales
B + C, exact£152Good intent, pricier clicks
B + C, phrase, –recipe£30Phrase reach brings junk the negative only half removes
B + C, –vegan£0The trap: no ads shown at all
B + C, broad, –recipe−£122Broad junk outweighs the sales
A + E, exact−£273"chocolate" spends the whole budget by day 6
A + D, broad, none−£441The worst: broad, busy and no negatives

Round 2 (18–34s reached)

TikTok CPM 70,000 · TikTok CPV 65,000 · Meta CPM 45,833 · YouTube CPV 40,625 · Meta CPV 39,722 · YouTube CPM 35,714 · TikTok CPC 19,444 · Meta CPC 18,333 · YouTube CPC 10,000 · Google Search (any) 5,000. The gap between CPC and CPM on the same platform is the lesson: paying per click for an awareness goal buys clicks, not eyeballs.

Round 3 (profit after ads, by chaos card)

Strategy + fixNo cardBotsTrackingRivalAI Overview
Target ROAS + page£279£245−£11£166£195
Max conversions + page£231£194−£39£109£162
Target ROAS + copy£204£174−£54£149£143
Max conversions + assets£129£98−£104£24£103
Manual CPC + page£108£16£108£6£75
Manual CPC + copy£40−£41£40−£9£28
Max clicks + page£34−£46£34−£55£24
Target ROAS + raise bids£25£4−£157−£46£18
Max clicks + raise bids−£166−£216−£166−£221−£116

When "Tracking broke" comes up, the manual bidder with a good landing page wins the round, which is the point of that card. Raising bids instead of fixing quality loses money with every strategy except Target ROAS, and even then it barely breaks even.

The screenshots below come from one four-team game: Truffle (B + E exact; TikTok CPM; Target ROAS + page), Fudge (C + E phrase –recipe; Google; Max clicks + raise bids), Praline (A + E exact; YouTube CPV; Max conversions + assets) and Ganache (B + C –vegan; Meta CPC; Manual + copy), with the AI Overview card. Final: Truffle 27, Praline 18, Fudge 18 (Praline ahead on round 3 profit), Ganache 15.

Timing

How long will it take?

Every time in this guide is a range. The low end is a tight run; the top end is what most classes take once teams discuss, decide and react to the results.

In class: 15 to 25 minutesSolo: at the player's own pace

What moves you up or down the range

  • How strictly you run the clock. Stopping each round when the timer ends keeps you near the low end; letting teams finish talking moves you to the top.
  • Number and size of teams. More teams means more picks to enter and more results to read out; bigger teams take longer to agree.
  • How much discussion you want. Quick calls, or asking each team to explain its choice before the reveal.
  • Getting started. Forming teams, explaining the rules and first-time players all take a few minutes.
  • Interaction in between. Questions, cheering at the results, the chaos card or a tech hiccup.
  • The debrief. One question, or a full discussion of what the game showed.
  • Your group. Familiarity with the topic, reading in a second language, and the room layout.

Short on time? Use the timers, keep to four teams or fewer, and take one debrief question. Want the richest discussion? Plan for the top of the range.

In class

Step by step

15 to 25 minutes with one debrief question. Short of time? Skip round 2: press Reveal the reach without entering any picks (every team scores the same), then Next round (10 to 15 minutes).

  1. 0

    Set the scene and form teams

    1 to 2 min
    The brief: In class or Solo, the Nibbo client card, three rounds and the points rule, and team names.
    The brief, the client and the winning rule on one screen.
    Click
    1. Set the number of teams with − and +; type team names if you like.
    2. Full screen, then Start round 1.
    Say
    "Each team is a PPC agency. Your client sells vegan dark chocolate online and nobody knows the name yet. You get £500 a round. Every click costs money; only buyers make money. Most points after three rounds takes the golden wrapper."
  2. 1

    Round 1: keywords, match type, negative

    3 to 6 min · timer 2:00
    Round 1: six keyword cards with weekly searches and CPC, three match types and three negative keywords.
    Six keywords, three match types and three negatives.
    The pick panel: a row for each team with keyword letters, match type and negative buttons.
    Tap each team's picks as spokespeople call them out.
    Click
    Start timer, tap in picks, read them back, then Reveal the money.
    Watch for
    • Someone picks "chocolate" (A) for its volume. Don't comment; the bar will.
    • "–vegan" sounds tidy. Ask the team that picked it why their ad never showed.
    Round 1 results: the winner’s name with shining stars, chocolate-bar profit bars, a reason for each team, and spend, clicks, orders and waste.
    The winner’s name shines with gold stars; each team gets a one-line reason.
  3. 2

    Round 2: platform and pay model

    2 to 4 min · timer 1:30
    Round 2: a Valentine's awareness brief with four platforms and three pay models (CPC, CPM, CPV).
    Now the goal is to be seen, not to sell.
    Say
    "New brief. Valentine's is coming and the goal is for as many 18 to 34 year olds as possible to see the video ad. Which platform, and how do you want to pay?"
    Watch for
    Teams that loved round 1 often stay on Google Search. The result explains why Search is small for awareness.
    Round 2 results: people aged 18–34 reached for each team, from 70,000 on TikTok CPM to 5,000 on Google Search.
    The same £500 reaches 5,000 or 70,000 people.
  4. 3

    Round 3: bid smart, then chaos

    2.5 to 5 min · timer 1:30
    Round 3 results with the chaos card: an AI Overview appears and teams lose 30% of clicks unless they added assets.
    The chaos card is drawn after everyone has chosen.
    Click
    Lock picks and draw a chaos card. Read the card aloud before scrolling to the results.
    Watch for
    Teams that picked "raise bids 15%" are surprised it loses. Ask: "What did you pay more for?"
  5. 4

    Podium, then switch the ads off

    1.5 to 3 min
    Final podium with gold, silver and bronze trophies, chocolate blocks and the winner’s name with shining stars.
    Melting chocolate, a fanfare, trophies and chocolate rain.
    Illustrative chart: paid search visits stop the week the ads are switched off; organic visits grow slowly and keep coming.
    "Now switch the ads off…": PPC is a tap, SEO is a tree.
    Say
    "Everything you just did stops the day the money stops. That's the difference from SEO, which you met in the last lecture."
On their own

Solo mode

Solo mode on a phone: the player taps their picks; three rival agencies have already decided.
In solo mode students play against three rival agencies. It works on a phone.

Share lab.yingwu.co.uk/hungry-for-clicks/#solo for homework or revision. The rivals always make the same choices: Broad Bros go big and hope, Safe & Sound play it safe, and Smart Cookies read the brief but aren't perfect. Finishing any game earns the "Hungry for clicks" badge; beating all three rivals in solo mode earns "Golden wrapper".

Model answers

Debrief questions

  1. Which single choice cost your team the most? Usually "chocolate" in round 1 (volume without intent), paying per click in round 2 (the wrong price for an awareness goal), or raising bids in round 3 (paying more for the same quality). Ask teams to name the choice and the metric that showed it.
  2. Where would SEO beat PPC for Nibbo, and where would PPC win? SEO wins for steady, long-term traffic on informational searches ("is dark chocolate vegan?") and once the brand is known. PPC wins for launches, seasonal pushes such as Valentine's, testing which keywords sell, and appearing for competitive commercial searches straight away. Most brands use both.
  3. If Nibbo gave you a real £500 tomorrow, what would you test first? Good answers set up conversion tracking first, start with brand and high-intent keywords, add negatives from the search terms report, fix the landing page, and only move to smart bidding once there is conversion data.
  4. Why did smart bidding lose when tracking broke? Smart bidding optimises towards the conversions it can see. If the tag stops firing it learns from wrong data. Check tracking before handing bidding to the algorithm.
  5. What do AI Overviews change? Fewer clicks for everyone on some searches, and new places for ads (above, below or inside AI Overviews). Richer ads, assets and strong landing pages matter more.
Keep on your phone

Cheat card: the button order

  1. Set teams, Full screen, Start round 1
  2. Start timer (2:00), tap picks, Reveal the money
  3. Next round, Start timer (1:30), tap platform and pay, Reveal the reach
  4. Next round, Start timer (1:30), tap strategy and fix, Lock picks and draw a chaos card
  5. Read the chaos card aloud; To the podium
  6. Now switch the ads off…; ask one debrief question
Sources

Behind the rules

A simplified classroom model: searches, prices and conversion rates are illustrative teaching devices. Nibbo is fictional.