2-day innovation sprint

The room in Toronto did not need another strategy offsite. Inkeros, founded in 2001, is a Canadian omnichannel sales and engagement company with 600-plus employees across Canada and the United States. The people in the chairs already knew the customer seams: sales, service, and the messy middle where a campaign promise meets an operations calendar. What they did not have was a format that would force a decision before the flight home.

Founded in 2001, Inkeros needed a format that would not vanish into a strategy offsite. We ran a two-day innovation sprint: problem framing on day one, prototypes and decision criteria on day two. The residue was a shortlist with owners. Topics in the room: Lean Startup T&E, workshop plus hackathon discipline, Design Thinking T&E.

The scene, without theatre

You do not walk into a 600-person sales-and-engagement business and ask them to "be more innovative." They are already innovating the way operators do: a workaround here, a campaign there, a vendor pilot that nobody killed. The Head of Operations, the sales directors, the CX or engagement leads, and a handful of builders already had opinions. Our job was to make those opinions testable.

Day one opened on a real account journey, not a warm-up. Where does a customer get a promise in one channel and a different truth in another? Where does a sales motion create work that service cannot see? Where does a US team and a Canadian team diverge on the same playbook? Those are Design Thinking questions with Lean honesty: if we cannot name the behaviour we will measure, we do not have a problem. We have a vibe.

The constraint owners sat in. In a commercial omnichannel shop the constraint is not APRA. It is brand claims, data use, channel economics, and the operations calendar. Treat those as users. Interview them. Write them on the brief. Otherwise you will prototype a campaign the floor cannot run.

The residue was a shortlist with owners. If your sprint produced a mural, you ran day one twice.

Day one: frame until it hurts

We did not generate 100 ideas. We generated a problem page.

• Customer words for the seam
• Operator words for the breakage
• Hypothesis (if we change X, Y moves)
• Thinnest test in 30 days
• Who could own it on Monday
• What would make us stop

Several candidate seams died before lunch because nobody could name an owner or a metric. That is a successful morning. Inkeros leads who live in revenue and retention already respect a kill. Give them the chance to use it.

Afternoon of day one was optioning, not visioning. Every option had to name a 30-day test: a script, a concierge, a landing page, a revised handoff checklist, a shadow process between sales and service. No platforms. No "we will stand up a lab."

Day two: prototype the decision

Teams built throwaway artefacts. Some were customer-facing. Some were internal punch lists (CX maps as operating lists, not murals). Decision criteria went on the wall in language a director could take to a Monday meeting: metric, date, dollars, stop.

The afternoon was a choice, not a showcase. Two ideas funded. The rest killed in public, with a sentence of learning. The person who could say no was in the chair. If they had sent a delegate, we would have had a demo day. We did not come for a demo day.

Hackathon energy showed up (it always does on day two). We kept it inside product-sprint rails: one problem per team, a hypothesis, a test path. Energy without a test path is a party. Inkeros did not need a party.

What travelled, and what we will not claim

We will not invent a revenue number. We will not quote a living executive. We will say what the format is for: a cross-border omnichannel operator that needs a shortlist with owners in 48 hours.

What travels to a peer in the US, to a retailer in Australia, to a telco engagement team in Singapore:

1. Frame on day one until seams have owners.
2. Prototype artefacts and decision criteria on day two.
3. Fund two. Kill the rest in the room.
4. Book the 30-day review before you leave.

If you are the operator

If you run a 400-to-800 person omnichannel shop in Canada or the US (or a sibling in AU/UK), do not book a venue until you have: a seam list from last month's real failures, a director who can fund two tests, and a risk or brand owner for claims. Then run two days. Circulate the shortlist the same afternoon. Delete the readout meeting.

Collective Campus will keep using Inkeros as a named case because the company is real, the footprint is CA/US, and the format left a residue. That is the bar.

Hour-by-hour run of show

A two-day innovation sprint at Inkeros only works if the clock is a product, not a vibe. Collective Campus ran the room that way. Inkeros, founded in 2001, is a Canadian omnichannel sales and engagement company with 600-plus people across Canada and the United States. The people in the chairs already knew the seams. They did not need a keynote. They needed a format that would force a shortlist with owners before anyone booked a flight home.

Steal the hours. Do not steal a mood board.

Day one, 08

to 09

. No icebreaker. No "what animal is your innovation style." Constraint owners sit: brand and claims, data use, channel economics, the operations calendar. The director who can fund two tests is in the chair or the sprint is a rehearsal. Inkeros-scale rooms fail when the weekend is energy and Monday is nobody's job. Name that person on the wall in the first half hour.

Day one, 09

to 09

. Seam list from last month's real failures, not a persona pack. Campaign promise in one channel, different truth in another. A sales motion that created work service could not see. A US playbook and a Canadian playbook that diverged on the same account. Those are the opening exhibits. If the list is empty, you booked the wrong week. Go get last month's tickets, chat transcripts, and the three accounts that made a director swear. Then start.

Day one, 09

to 11

. Walk the omnichannel handoffs. Not a journey mural. A walk. One account, one promise, one breakage. Collective Campus facilitated. Inkeros operators narrated. The walk is the Design Thinking move with Lean honesty: if we cannot name the behaviour we will measure, we do not have a problem. We have a vibe. Detail on the walk sits in the next section. Protect this block. It is the only research the room needs if last month's failures are real.

Day one, 11

to 12

. Kill. Several candidate seams die before lunch because nobody can name an owner or a metric. That is a successful morning. Inkeros leads who live in revenue and retention already respect a kill. Give them the chance to use it before sandwiches. What dies, and why, is its own section below. Do not skip the funeral. A seam that limps into the afternoon will steal prototype hours from a seam that can live.

Day one, 12

to 13

. Lunch is not a networking slot. The named director and the constraint owners eat with the facilitators. The question is crude: which two seams still have an owner if we leave tonight? If the answer is "we will see after optioning," the morning failed. Optioning does not create owners. Owners walk in or they do not.

Day one, 13

to 15

. Optioning, not visioning. Every option names a 30-day test: a script, a concierge, a landing page, a revised handoff checklist, a shadow process between sales and service. No platforms. No "we will stand up a lab." No vendor that needs a statement of work before a customer sentence exists. Lean Startup in this block means a hypothesis (if we change X, Y moves) plus a thinnest test. Design Thinking in this block means the customer language from the morning walk stays on the option card. If the option cannot be said in the customer's words, it is a system story. System stories wait.

Day one, 15

to 17

. Problem pages. One page per surviving seam. Customer words. Operator words. Hypothesis. Thinnest test in 30 days. Monday owner. Kill criterion. Circulate the pages the same afternoon. Delete the readout meeting that would have restated them on Wednesday.

Day one, 17

to 17

. Residue check. If you do not have two to four problem pages with named owners, you do not have a day two. You have a day one to repeat. Collective Campus will cancel a pretty afternoon of prototyping before we run a hackathon in costume.

Day two, 08

to 09

. Recap the shortlist in ten minutes. The director restates which seams still have money. If they send a delegate who cannot kill, you scheduled a demo day. We did not come for a demo day.

Day two, 09

to 12

. Prototype the artefacts and the decision criteria in parallel. Some artefacts are customer-facing: a script, a page, a revised first email, a chat-to-voice handoff card. Some are internal punch lists: the CX map as an operating list, not a mural. Decision criteria go on the wall in language a director can take to a Monday meeting: metric, date, dollars, stop. Hackathon energy will show up. It always does on day two. Keep it inside product-sprint rails: one problem per team, a hypothesis, a test path. Energy without a test path is a party. Inkeros did not need a party.

Day two, 12

to 12

. Lunch. Same rule as day one. The person who can say no does not disappear into email.

Day two, 12

to 14

. Tighten. Expose each artefact to someone who was not on the build team: a floor lead, a sales manager, a service coach. Repair what they cannot run on a wet Tuesday. If they need the facilitator to decode the card, rewrite the card.

Day two, 14

to 16

. Choice, not a showcase. Two ideas funded. The rest killed in public, with a sentence of learning. The person who can say no is in the chair. Applause is optional. The shortlist is not.

Day two, 16

to 16

. Monday budget line. Named owners. Calendar hold for the 30-day review, booked before anyone stands up. What "budget line" means, and what it does not mean, is its own section. Do not leave this as a vibe.

Day two, 16

to 17

. Circulate the shortlist. Same afternoon. If a slide deck is required for a steering pack, it is an appendix to the shortlist, not the other way around.

That is the run of show Collective Campus delivered. It travels. It is not a product endorsement of Inkeros. It is a format with a clock.

How the omnichannel handoffs were walked

Omnichannel operators live or die in handoffs. A 600-person sales-and-engagement business already knows this. The sprint does not discover the idea of a seam. It walks one.

We did not start with a persona named Omni Olivia. We started with a real account journey Inkeros people could recognise without a caption. Where does a customer get a promise in email or chat and a different truth on the phone? Where does a sales motion create a task that service cannot see in the same shift? Where does a US team and a Canadian team diverge on the same playbook, same brand, different clock? Those are the walks.

A walk has four beats, and they fit on one page.

The promise. What did we say, in which channel, in whose voice? Write the sentence. Not the campaign name. The sentence the customer could forward to a colleague.

The breakage. Where did the next human or the next system fail to inherit that sentence? Chat to voice. Voice to a specialist. Sales to service. Campaign ops to the floor calendar. Canada to the United States. The breakage is usually a missing field, a missing owner, or a missing hour, not a missing strategy.

The operator sentence. What does the floor actually do when the breakage hits? Workaround, queue, apology, second contact. If you cannot get that sentence from a coach or a team lead, you are still in marketing. Sit with the coach.

The measure. Repeat contact, time to a true answer, a sales-created task that service closed without seeing the promise, a cross-border play that needed a translator inside the company. If the measure needs a data-warehouse project, you picked a vanity. Pick a count a team lead already argues about.

Collective Campus walked those beats with Inkeros operators in the room. That mix is the only one that produces a residue. A facilitator alone will mural the handoff. A marketer alone will campaign the handoff. An engineer alone will platform the handoff. An operator plus a director plus a constraint owner will name the checklist, the script, or the shadow process that can be tested in 30 days.

Cross-border is a special case, not a subplot. Inkeros runs Canada and the United States. The same playbook can be legal in one and loud in the other. Claims, data use, and the operations calendar diverge even when the brand does not. Walk one account that crossed the line. Write the two truths on the problem page. A peer in Australia and New Zealand (a Woolworths-shaped retailer, a Telstra-shaped care team) already knows this physics between states or across the Tasman. A peer in the United Kingdom (a BT-shaped or Tesco-shaped engagement team) already knows it between brands in a group. Steal the walk, not a global operating model poster.

What we will not do in the walk: invent a quote from a living Inkeros executive, invent a conversion rate on a channel, or pretend a CRM screenshot is a journey. The walk is a conversation with the last real failure. If that failure is not in the room, you are doing theatre with better snacks.

What died before lunch, and why that was the morning

Several candidate seams died before lunch on day one because nobody could name an owner or a metric. That sentence is already the case. This section is the method, so a peer can run the same funeral.

Things that die in an Inkeros-scale omnichannel room, and should:

The platform that would "fix handoffs." If the owner is a vendor and the metric is "adoption," it is not a seam. It is a procurement. Platforms might be later. They are not a 30-day test. Kill them before lunch so they do not eat the prototype afternoon.

The brand campaign with no floor path. A promise the operations calendar cannot keep is a lie with a media plan. Brand and claims sit in the room so this dies early. Inkeros people who have already cleaned up a campaign-created queue do not need a lecture. They need permission to say no in front of a director.

The lab. "We should stand up an innovation function" is not a customer problem. It is a career problem. Kill it. The residue of this sprint is a shortlist, not an org chart.

The cross-border unification programme. Canada and the United States will not share one playbook because a workshop asked nicely. If no director owns a 30-day test on one diverging play, the unification story dies. A later programme can earn its way. It does not earn the afternoon.

The idea with energy and no kill. Someone will love a concept. Ask for the metric and the name. If both are missing, the concept had a good morning and a closed casket. Inkeros leads who live in revenue and retention already know how to do this in a pipeline review. Use that muscle.

The persona refresh. If last month's failures are on the table, you do not need new personas. You need owners. Personas that cannot name a breakage from last month die before lunch.

The funeral is public and short. One sentence of learning per kill. "No owner." "No metric a team lead already tracks." "Needs production on day one and risk is not in the chair." "Campaign without a floor path." Write it on the problem wall so it cannot be resurrected as a side conversation at 15

.

This is Lean Startup honesty inside a Design Thinking frame. Hypothesis plus customer language, plus a kill a director will sign. Collective Campus will keep killing before lunch. That is the work. The mural was never the work.

A peer running this in Sydney or London will feel the same deaths. Woolworths-shaped and Coles-shaped retail engagement will try to keep a campaign. Telstra-shaped care will try to keep a platform. A UK engagement team inside BT, Sky, or a major grocer will try to keep a lab. Let them try. Then ask for the owner and the metric. The morning is for that question, not for ideation volume.

The Monday budget line

A two-day sprint is justified only if Monday has a budget line. Not a new cost centre. Not an innovation envelope that lives in a lab. A line a director can sign for two 30-day tests.

Here is what that line contains. Steal the list.

Two funded tests. Not a portfolio of twelve. Two. Each has an artefact (script, concierge, page, checklist, shadow), a metric, a named owner, and a kill date. The rest died in the room. If Monday's email says "we will socialise the full set," you ran a demo day and then got scared.

Dollars the owner already controls. Travel leftover, a campaign scrap, a contractor week, a floor coach's time. We will not invent an Inkeros figure. We will say the money has to be real enough that a director can point at it without a business case. If the test needs a business case, it is not a 30-day test. It is a project wearing a sprint hoodie.

A calendar hold for the 30-day review. Booked before people stand up on day two. Same director. Same constraint owners. Same two owners of the tests. If aftercare has no home, you will get a sequel workshop. Sequels are how theatre bills you twice.

A stop that is already written. "We will stop if we cannot get the floor to run the checklist without a facilitator." "We will stop if repeat contact on the walked account type does not move." "We will stop if brand will not let us say the sentence the test needs." Boards and directors relax when they can see the stop. Inkeros-scale commercial rooms already live this way in pipeline. Steal that for experiments.

A named person who will write the funeral or the keep. Residue is a page, not a feeling. The owner writes it. Collective Campus can facilitate the review. We cannot own Inkeros's Monday. Position B: we delivered the sprint. Monday is theirs.

What the budget line is not: a licence to stand up a lab, a vendor retainer, a platform proof of concept, or a cross-border transformation office. Those might be later, if two tests earn them. They are not the residue of 48 hours.

If you cannot name the line before you book the venue, do not book the venue. Run a 90-minute framing call instead. Collective Campus would rather lose a room than run a pretty weekend that dies on Monday.

How a peer in Australia or the United Kingdom steals the format

The format travels. The letterhead changes. The physics do not.

If you run a 400-to-800 person omnichannel shop in Australia, start with last month's real failures in a Woolworths-shaped retail network, a Telstra-shaped care and field operation, or a specialist engagement team that already lives in campaign-to-floor breakage. Book a director who can fund two tests. Book brand or claims. Book the floor coach. Run the hours above. Walk one substitute failure, one bill-shock call, one promise a store or a care bay could not keep. Kill before lunch. Prototype artefacts and decision criteria on day two. Circulate the shortlist the same afternoon. The Tasman fork (Australia and New Zealand) is your version of Inkeros's Canada and United States fork. Walk one account that crossed it.

If you run the sibling shop in the United Kingdom, start with a BT-shaped or Sky-shaped care motion, a Tesco-shaped or John Lewis-shaped retail engagement team, or a mid-market sales shop that already lost a week to a campaign the floor could not run. Ofgem-shaped and FCA-shaped claims rules are your constraint owners. They are not villains. They are design inputs. Put them in at 08

on day one, the way Inkeros put brand, data, and the operations calendar in the chair. Walk one handoff. Kill the platform story before lunch. Fund two tests a director can sign without a steering saga.

Hong Kong and Singapore peers (a Singtel-shaped care lead, an HSBC Hong Kong-shaped or DBS-shaped service desk that keeps paying for campaign-created rework) steal the same hours. The constraint is conduct and the operations calendar, not a maple leaf. A Canadian or US peer already in Inkeros's neighbourhood does not get a different religion. They get the same shortlist rule: two funded, the rest killed in the room, a 30-day review booked before wheels up.

What you copy is the run of show, the walk, the funeral, and the Monday line. What you do not copy is a claim about Inkeros's revenue, a quote we did not hear, or a conversion rate we will not invent. Collective Campus will keep using Inkeros as a named case because the company is real, the footprint is Canada and the United States, and the format left a residue. A peer steals the format. They do not steal a fabricated scoreboard.

A practical booking note:

We will run two days. Day one frames until seams have owners. Day two prototypes artefacts and decision criteria. We will walk one real omnichannel handoff. We will kill before lunch anything without an owner or a metric. We will fund two 30-day tests with a Monday budget line and a review already on the calendar. We will not stand up a lab. If you cannot name the director and last month's failures, we will not book the venue.

What we will not claim about Inkeros

Position B, said in plain words. Collective Campus delivered a two-day innovation sprint with Inkeros. That is why we can describe the room, the hours, the walk, the funeral, and the Monday line. It is not an endorsement of any Inkeros product, platform, or campaign. It is not a licence to invent their numbers.

We will not invent a revenue lift. We will not invent a conversion rate on a channel or a page. We will not quote a living executive as if they sat for a profile. We will not claim a firm-wide transformation of a 600-person company in 48 hours. We will not claim the two funded tests "won." We will say they had owners, kill dates, and a review on the calendar. That is the residue. That is the case.

We will not claim Lean Startup or Design Thinking as a badge Inkeros now wears. Those were topics in the room: hypothesis plus customer language, plus a kill a director will sign. Plenty of companies buy the badge and skip the kill. We will not help Inkeros, or a peer, launder a mural into a method.

We will not claim the format is unique to Canada. It is not. It is honest in any omnichannel shop in our geos: Australia, New Zealand, Hong Kong, Singapore, the United States, the United Kingdom, and Canada. We name Inkeros because the client is real and the cross-border footprint is the story. Naming is how target operators find the piece. Fabrication is how they bounce.

If a case study needs a percentage to be interesting, it is not a case. It is an ad. This is a field note on a format. Steal the hours.